feat: add policy evaluation heatmap to SRE dashboard (#309) by zeel2104 · Pull Request #326 · microsoft/agent-governance-toolkit

zeel2104 · 2026-03-21T15:35:07Z

Description

Add a Policy Evaluation Heatmap widget to the Agent SRE observability dashboard.
The heatmap visualises governance metric density over time, making it easy to spot
patterns like "Agent X hits content policy limits every Tuesday at 3pm".

Features added:

🔥 New "Policy Heatmap" tab in examples/dashboard/app.py
Toggle between Agent × Time and Policy × Time heatmap axes
Toggle color metric between Evaluation Count and Violation Count
Hover tooltip shows exact count per cell
KPI strip: total evaluations, total violations, violation rate, peak activity
Evaluation & violation distribution bar charts
Summary table sorted by violations with Peak Bucket column
Data shape matches governance metrics from the OpenTelemetry pipeline

Type of Change

Bug fix (non-breaking change that fixes an issue)
New feature (non-breaking change that adds functionality)
Breaking change (fix or feature that would cause existing functionality to change)
Documentation update
Maintenance (dependency updates, CI/CD, refactoring)
Security fix

kernel

Package(s) Affected

Checklist

My code follows the project style guidelines (ruff check)
I have added Screenshots that prove my fix/feature works
All new and existing tests pass (pytest)
I have updated documentation as needed
I have signed the Microsoft CLA

Related Issues

Closes #309

github-actions · 2026-03-21T15:35:19Z

Welcome to the Agent Governance Toolkit! Thanks for your first pull request.
Please ensure tests pass, code follows style (ruff check), and you have signed the CLA.
See our Contributing Guide.

github-actions · 2026-03-21T15:35:29Z

🤖 AI Agent: contributor-guide

Welcome 🎉

Hi @first-time-contributor! 👋 Welcome to the microsoft/agent-governance-toolkit project, and thank you for taking the time to contribute! We’re thrilled to have you here. 😊

Your pull request to add a Policy Evaluation Heatmap to the SRE dashboard is a fantastic addition. Visualizing governance metrics in this way will undoubtedly help users identify patterns and optimize their agent configurations. Great work!

What You Did Well ✅

Clear Description: Your PR description is detailed and well-structured. It clearly outlines the new feature, its purpose, and the changes made. The inclusion of screenshots is a great touch — it makes it easy to visualize the feature without running the code.
Thoughtful Design: The heatmap functionality is well-thought-out, with useful toggles for axes and color metrics. The addition of KPIs, bar charts, and a summary table makes the feature comprehensive and user-friendly.
Code Quality: Your code is clean, well-documented, and follows Python best practices. The use of helper functions like gen_policy_heatmap_data is excellent for readability and maintainability.
Testing and Documentation: It's great to see that you've run ruff for linting, added screenshots, and updated the documentation. These are essential steps for a successful contribution.

Suggestions for Improvement ✨

Tests: While your feature looks great, I noticed that there are no new tests included in this PR. Since this is a new feature, it would be ideal to add tests to ensure the functionality works as expected and remains stable over time.
- Project Convention: Tests for the agent-sre package should go in the packages/agent-sre/tests/ directory. You can create a new test file, such as test_policy_heatmap.py, and include unit tests for the gen_policy_heatmap_data function and any other critical logic.
- Resource: Check out our CONTRIBUTING.md for more details on testing.
Conventional Commit Message: Your commit message is close to the format we use, but it could be slightly refined. For example:
```
feat(agent-sre): add policy evaluation heatmap to SRE dashboard
```
This format specifies the scope (agent-sre) and aligns with our conventional commits guidelines.
Security Considerations: Since this feature involves governance metrics, it’s important to ensure that no sensitive data is exposed in the heatmap or summary table. For example:
- Are there any privacy concerns with displaying agent or policy names?
- Could the data in the heatmap or tooltips reveal sensitive information?
- If applicable, consider adding a note in the documentation about how this feature handles sensitive data.
Code Style: While your code is generally clean, running ruff on the new code might reveal some minor improvements. For example:
- Ensure consistent line breaks and spacing.
- Check for unused imports or variables.

Next Steps 🚀

Add Tests: Please add tests for the new functionality in packages/agent-sre/tests/. This will help ensure the feature is robust and maintainable.
Update Commit Message: Update your commit message to follow the conventional commits format.
Run Ruff: Double-check your code with ruff to ensure it adheres to our linting rules.
Security Review: Confirm that the feature does not expose sensitive data and adheres to our security guidelines.

Once you’ve made these updates, push your changes to this branch, and we’ll review them again. If you have any questions or need help with anything, feel free to ask — we’re here to support you!

Helpful Resources 📚

CONTRIBUTING.md: Guidelines for contributing to this project.
QUICKSTART.md: A quick guide to getting started with the project.
Conventional Commits: https://www.conventionalcommits.org/

Thank you again for your contribution! We’re excited to see this feature come to life. 😊

github-actions

🤖 AI Agent: code-reviewer

Review Summary

This pull request introduces a Policy Evaluation Heatmap widget to the Agent SRE observability dashboard. The feature is well-documented and provides a visual representation of governance metric density over time, which is a valuable addition for monitoring and identifying patterns in agent behavior. The implementation includes a heatmap, KPI metrics, bar charts, and a summary table, all of which are useful for analyzing policy evaluations and violations.

However, there are a few areas of concern and improvement that need to be addressed before merging this PR.

🔴 CRITICAL

Potential Data Leakage in Heatmap Tooltip
- Issue: The hover tooltip in the heatmap displays detailed information about agents and policies, including evaluation and violation counts. If this dashboard is exposed to unauthorized users, it could lead to sensitive information leakage.
- Recommendation: Implement access control to ensure that only authorized users can view the dashboard. Additionally, consider obfuscating sensitive agent or policy identifiers in the tooltip if the dashboard is accessible to a broader audience.
Random Data Generation for Simulations
- Issue: The gen_policy_heatmap_data function uses random number generation (rng.uniform, rng.integers) to simulate data. While this is acceptable for testing, it could lead to inconsistencies in the dashboard's behavior if the random seed is not fixed.
- Recommendation: Use a fixed random seed for reproducibility during testing. For example, rng = np.random.default_rng(seed=42).
Time Zone Handling
- Issue: The gen_policy_heatmap_data function uses datetime.now(tz=dt.timezone.utc) to calculate time buckets. If the dashboard is used across multiple time zones, this could lead to incorrect or confusing time labels.
- Recommendation: Allow the user to specify the desired time zone for the heatmap or ensure that the time zone is clearly displayed on the dashboard.

🟡 WARNING

Backward Compatibility
- Issue: The addition of the "Policy Heatmap" tab modifies the examples/dashboard/app.py file, which might affect users who rely on the existing dashboard layout.
- Recommendation: Confirm whether this change is backward-compatible with existing workflows. If not, document the changes clearly in the release notes and consider providing a migration guide.

💡 SUGGESTIONS

Unit Tests for gen_policy_heatmap_data
- Observation: The gen_policy_heatmap_data function is a critical part of this feature, but it lacks unit tests.
- Recommendation: Add unit tests to validate the correctness of the generated data, especially for edge cases like:
  - Empty selected_agents or POLICIES.
  - Very small or very large time_range values.
  - Handling of missing or incomplete data.
Type Annotations
- Observation: The gen_policy_heatmap_data function uses type annotations, but some parts could be more explicit. For example:
```
view: str = "agent_x_time"
```
  could be replaced with:
```
view: Literal["agent_x_time", "policy_x_time"] = "agent_x_time"
```
- Recommendation: Use more specific type annotations (e.g., Literal) where applicable to improve type safety and code readability.
Error Handling for DataFrame Operations
- Observation: The function assumes that eval_df and viol_df will always contain valid data. However, if the input data is malformed or empty, operations like nlargest or pivot could raise exceptions.
- Recommendation: Add error handling to gracefully handle cases where the input data is empty or malformed. For example:
```
if eval_df.empty or viol_df.empty:
    st.warning("No data available for the selected time range or filters.")
    return pd.DataFrame(), pd.DataFrame(), pd.DataFrame()
```
Performance Optimization
- Observation: The gen_policy_heatmap_data function generates data for all agents and policies, even if only a subset is selected for display.
- Recommendation: Optimize the function to generate data only for the selected agents or policies to improve performance, especially for large datasets.
Documentation
- Observation: While the feature is well-documented in the PR description, there is no mention of this new feature in the user-facing documentation.
- Recommendation: Update the user documentation to include details about the new "Policy Heatmap" tab, its features, and how to use it effectively.
Accessibility
- Observation: The heatmap and bar charts rely heavily on color to convey information, which may not be accessible to colorblind users.
- Recommendation: Add patterns or textures to the heatmap and bar charts to make them more accessible. Alternatively, provide a colorblind-friendly color palette.
Dynamic Bucket Width
- Observation: The bucket width is determined based on the time_range, but the logic is hardcoded and may not cover all use cases.
- Recommendation: Allow users to customize the bucket width via a dropdown or slider in the dashboard.
Code Style
- Observation: The gen_policy_heatmap_data function has some repeated logic for generating evaluation and violation data for agents and policies.
- Recommendation: Refactor the function to reduce code duplication. For example, you could extract the common logic into a helper function.

Conclusion

The new feature is a valuable addition to the Agent SRE observability dashboard, but the identified issues, especially the critical ones, need to be addressed to ensure security, reliability, and usability. Once these concerns are resolved, the PR can be approved for merging. Let me know if you need further clarification or assistance!

github-actions · 2026-03-21T15:35:33Z

🤖 AI Agent: security-scanner

Security Review of PR #309: Adding Policy Evaluation Heatmap to SRE Dashboard

This PR introduces a new feature to the Agent SRE observability dashboard: a Policy Evaluation Heatmap. While the feature itself is primarily for visualization and monitoring, it interacts with governance metrics and OpenTelemetry data, which are critical components of the agent governance toolkit. Below is the security analysis based on the specified criteria:

1. Prompt Injection Defense Bypass

Risk: 🔵 LOW

Analysis: The feature does not directly handle user-provided input that could be used for prompt injection. The heatmap data is generated from OpenTelemetry metrics or simulated data, and there is no evidence of user-controlled input being passed to the policy evaluation logic.
Recommendation: No immediate action required. However, ensure that any future integration with user-provided inputs (e.g., filtering or custom queries) includes proper sanitization and validation.

2. Policy Engine Circumvention

Risk: 🔵 LOW

Analysis: The PR does not modify the core policy engine or its enforcement mechanisms. The heatmap is purely a visualization tool and does not impact the actual evaluation or enforcement of policies.
Recommendation: No immediate action required. Ensure that future changes to the dashboard do not inadvertently expose mechanisms to bypass policy enforcement.

3. Trust Chain Weaknesses

Risk: 🔵 LOW

Analysis: The heatmap relies on OpenTelemetry data, but there is no evidence of changes to the trust chain or certificate validation mechanisms in this PR. The data is simulated for demonstration purposes.
Recommendation: Ensure that the OpenTelemetry pipeline is securely configured and that data integrity is verified in production environments.

4. Credential Exposure

Risk: 🔵 LOW

Analysis: There is no evidence of sensitive credentials being exposed in the code. The PR does not include any hardcoded secrets, and the data used is either simulated or sourced from a secure OpenTelemetry pipeline.
Recommendation: Continue to monitor for any accidental inclusion of sensitive information in future changes.

5. Sandbox Escape

Risk: 🔵 LOW

Analysis: The feature is a visualization tool and does not execute untrusted code or interact with the underlying system in a way that could lead to a sandbox escape.
Recommendation: No immediate action required.

6. Deserialization Attacks

Risk: 🟡 MEDIUM

Analysis: The gen_policy_heatmap_data function uses pandas.DataFrame to process data. While the current implementation generates simulated data, in a production environment, this function will likely process real OpenTelemetry data. If this data is not properly validated or sanitized, it could be a vector for deserialization attacks or other data integrity issues.
Recommendation: Add validation and sanitization steps for any data ingested from external sources (e.g., OpenTelemetry). Ensure that the data conforms to expected formats and does not contain malicious payloads.

7. Race Conditions

Risk: 🔵 LOW

Analysis: The feature does not introduce any concurrent operations or time-of-check-to-time-of-use (TOCTOU) vulnerabilities. The data is processed and visualized in a single-threaded manner.
Recommendation: No immediate action required.

8. Supply Chain

Risk: 🟠 HIGH

Analysis: The PR does not explicitly list any new dependencies. However, it uses pandas and plotly, which are external libraries. If these libraries are not pinned to specific versions in the requirements.txt or equivalent file, there is a risk of dependency confusion or malicious package injection.
Recommendation: Ensure that all dependencies are pinned to specific versions in the dependency management file. Regularly audit dependencies for known vulnerabilities using tools like pip-audit or safety.

Additional Observations

Random Number Generation: The gen_policy_heatmap_data function uses rng.uniform and rng.integers for generating simulated data. While this is acceptable for testing, ensure that in production, the data is sourced securely from the OpenTelemetry pipeline and not generated randomly.
Data Exposure: The heatmap and summary table display detailed governance metrics. While this is intended for observability, ensure that access to the dashboard is restricted to authorized users only. Sensitive governance metrics should not be exposed to unauthorized individuals.

Summary of Findings

Issue	Risk	Description	Recommendation
Prompt Injection Defense Bypass	🔵 LOW	No user-controlled input is processed in this PR.	No action required.
Policy Engine Circumvention	🔵 LOW	No changes to the policy engine or enforcement mechanisms.	No action required.
Trust Chain Weaknesses	🔵 LOW	No changes to trust chain or certificate validation.	Ensure OpenTelemetry pipeline is securely configured.
Credential Exposure	🔵 LOW	No sensitive credentials are exposed.	No action required.
Sandbox Escape	🔵 LOW	No untrusted code execution or system-level interactions.	No action required.
Deserialization Attacks	🟡 MEDIUM	Potential risk if OpenTelemetry data is not validated/sanitized in production.	Add validation and sanitization for ingested data.
Race Conditions	🔵 LOW	No concurrent operations or TOCTOU vulnerabilities.	No action required.
Supply Chain	🟠 HIGH	Dependencies (`pandas`, `plotly`) are not pinned to specific versions.	Pin dependencies to specific versions and audit them for vulnerabilities.

Final Recommendation

Approve with Changes: The feature itself does not introduce critical security risks, but there are medium and high risks related to deserialization and supply chain vulnerabilities. Address these issues before merging:
1. Add data validation and sanitization for OpenTelemetry data in production.
2. Pin all dependencies to specific versions and perform a security audit on them.

zeel2104 · 2026-03-21T15:39:49Z

@microsoft-github-policy-service agree

…icrosoft#326)

#327) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…ent-os replaced (#328) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

* docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…summary (#345) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(ci): add PR review orchestrator — collapse agents, post unified summary - Wrap individual agent comments in <details> tags (collapsed by default) - Make agent comments idempotent (update on re-push, don't duplicate) - Add ai-pr-summary.yml workflow that posts one clean verdict table - Summary uses HTML marker for upsert behavior Contributors now see ONE summary table instead of 5-7 separate bot comments. Individual agent reports are preserved but collapsed for reference. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…LICENSE (#344) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix: fill community PR gaps — bare excepts, print→logging, py.typed, LICENSE files Address remaining items from issues #330, #331, #332, #334 that community PRs partially covered. - Fix bare except blocks in time_travel_debugger.py, demo_client.py - Replace print() with logging in integrity.py, promotion.py, verify.py - Add py.typed markers for PEP 561 compliance - Copy LICENSE to all subpackages Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…ion (#346) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(compliance): add compliance_grade() method to GovernanceAttestation Adds a letter-grade (A-F) method based on OWASP ASI control coverage percentage. Updates README with usage example. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix: broken markdown links and lychee exclusions - Fix stargazers badge URL in agentmesh-integrations/README.md - Fix internal links to agent-os and agent-sre integrations - Add lychee exclusions for nuget (unpublished), nist.gov (flaky), stargazers paths, and not-yet-created compliance docs Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * ci: make link checker non-blocking until pre-existing links are fixed (#320) There are 20+ pre-existing broken links across docs/, proposals/, and tutorials/ that predate this PR. Setting fail: false so the check reports but doesn't block merges. Issue #320 tracks the full link cleanup. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…it CI (#350) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…icrosoft#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (microsoft#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes microsoft#316 * ci: add markdown link checker workflow (microsoft#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (microsoft#309) (microsoft#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (microsoft#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR microsoft#325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

* docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

* docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: refresh proposal status fields to reflect current state Update Status fields across all 8 proposal documents: - DIFY: Shipped (PR #2060, live on Marketplace) - GITHUB-COPILOT: Shipped (all 3 PRs merged) - MAF: Implemented Level 1 (18 tests, awaiting MAF team) - MCP: Partially Shipped (npm + Glama, registry pending) - OPENLIT: Implemented (PR #1062 under review) - CSA-ATF: Active (ATF author engaged, conformance spec published) - AAIF: Paused (re-submit after public release) - GOOGLE-ADK: Implemented (GovernanceAdapter shipped) Added Progress section to CSA-ATF-PROPOSAL.md with timeline. * docs: create 5 new proposals, update proposals index New proposals for shipped/active engagements: - HAYSTACK-INTEGRATION-PROPOSAL.md (shipped) - ORACLE-AGENTSPEC-PROPOSAL.md (active engagement) - STRIPE-MPP-PROPOSAL.md (planned) - NEXUS-TRUST-EXCHANGE-PROPOSAL.md (pre-alpha) - A2A-TRUST-EXTENSIONS-PROPOSAL.md (adapter shipped) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): dependency confusion in quickstart.ps1, bare excepts, MD5 deprecation CRITICAL: quickstart.ps1 referenced unregistered 'agent-os' PyPI name instead of 'agent-os-kernel'. Fixed to prevent supply chain attack. HIGH: 4 bare except: blocks in graph_debugger.py (2 duplicate files) replaced with specific exception types + logging. MEDIUM: MD5 and SHA1 deprecated with warning in text_tool.py hash algorithm selection. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…[MSRC-111178] (#353) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): harden pull_request_target workflows against fork RCE [MSRC-111178] Defense-in-depth hardening for all 6 AI agent workflows: - Explicit ref: base.sha on all checkout steps (was implicit default) - persist-credentials: false on all checkouts - Fork safety notices - Shallow clone (fetch-depth: 1) where full history not needed - Security header comments referencing MSRC case The core vulnerability (ref: head.sha checkout) was already removed in PR #303. These changes add belt-and-suspenders protection. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…to fallback (#354) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): add SECURITY.md to all packages, harden langchain crypto fallback - Add SECURITY.md to 29 packages (9 main + 20 integrations) linking to root security reporting guidance - Harden langchain-agentmesh crypto fallback: emit SecurityWarning when cryptography package unavailable (simulation mode) - Add explicit SECURITY WARNING comments on fallback code paths Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…rmissions (#355) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(deps): bump cryptography>=45.0.3, migrate PyPDF2→pypdf, scope workflow permissions Vulnerability fixes: - cryptography: bump minimum to >=45.0.3 (fixes GHSA-3ww4, GHSA-9v9h, GHSA-h4gh, GHSA-jm77, GHSA-r6ph, GHSA-v8gr) - PyPDF2: migrate to pypdf>=6.8.0 (fixes GHSA-4vvm, GHSA-hqmh, GHSA-qpxp) - nltk: add upper bound <4.0 (3.9.3 already patched) - black: bump to >=25.1.0 (fixes GHSA-3936) Workflow permissions: - ai-release-notes.yml: move contents:write to job level - ai-spec-drafter.yml: move contents:write to job level Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

* docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(scorecard): remove explicit ref in pull_request_target, add sigstore, pin CI deps Dangerous-Workflow: Remove ref: base.sha from pull_request_target workflows — default checkout IS the base branch, explicit ref triggers Scorecard false positive. Security preserved via persist-credentials: false + MSRC comments. Signed-Releases: Add sigstore/gh-action-sigstore-python to publish workflow for Python package signing alongside existing attest-build-provenance. Pinned-Dependencies: Pin pip install versions in CI workflows: - ci.yml: pytest==8.4.1, pytest-asyncio==0.26.0 - benchmarks.yml: add --no-cache-dir - security-scan.yml: pyyaml==6.0.2 - weekly-security-audit.yml: pyyaml==6.0.2 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…359) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes #316 * ci: add markdown link checker workflow (#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (#309) (#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR #325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(esrp): configure ADO pipelines with Client ID and team contacts Fill ESRP pipeline placeholders with: - Client ID: a458522c-0359-4e92-9887-5fee1607c0c7 - Service connection: agt-esrp-release - Owners/Approvers: agt@microsoft.com TODO remaining (this week): - Key Vault name (pending PRSS cert generation) - Auth/Sign cert names (pending SAW access) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

…rosoft#356) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (microsoft#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes microsoft#316 * ci: add markdown link checker workflow (microsoft#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (microsoft#309) (microsoft#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (microsoft#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR microsoft#325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * feat(security): add dependency confusion pre-commit hook + weekly audit CI - scripts/check_dependency_confusion.py: Pre-commit hook that scans for pip install commands referencing unregistered PyPI packages. Maintains an allowlist of known registered packages. - .github/workflows/weekly-security-audit.yml: Weekly CI job running dependency confusion scan, security skills scan, and weak crypto check. Uploads reports as artifacts with 90-day retention. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(security): proactive audit — dependency confusion + MD5→SHA-256 (microsoft#349) * docs: add testing guide for external testers and customers Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add regulatory alignment table and Purview positioning to README Add EU AI Act, Colorado AI Act, and GPAI obligations timeline with AGT coverage mapping. Reference Microsoft Purview DSPM for AI as complementary data governance layer. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(ci): restore read-all at workflow level for Scorecard verification The Scorecard API rejects workflows with write permissions at the workflow level. id-token: write and security-events: write must be scoped to the job level only. Restores permissions: read-all at workflow level while keeping job-level write permissions intact. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: add comprehensive docstrings to mcp_adapter.py classes (microsoft#324) Add Google-style docstrings with Args, Returns, Raises, Attributes, and Example sections to MCPMessageType, MCPAdapter, and MCPServer classes. Also enhances docstrings for key methods including handle_message, _handle_tools_call, _handle_resources_read, and _map_tool_to_action. Fixes microsoft#316 * ci: add markdown link checker workflow (microsoft#323) Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> * feat: add policy evaluation heatmap to SRE dashboard (microsoft#309) (microsoft#326) * fix: remove unregistered PyPI packages from notebooks and requirements (dependency confusion) (microsoft#325) - Replace !pip install agent-os with !pip install -e ../.. in all 6 notebooks; agent-os is not on PyPI and installing it from PyPI is a dependency confusion vector - Replace zendesk-sdk/freshdesk-sdk with zenpy/freshdesk (the real published SDKs) in customer-service/requirements.txt - Remove hashlib-compat from healthcare-hipaa/requirements.txt; hashlib is stdlib and hashlib-compat is not a real PyPI package * fix(security): complete dependency confusion fix — replace all pip install agent-os with agent-os-kernel Replace all remaining instances of `pip install agent-os` (unregistered on PyPI) with `pip install agent-os-kernel` (the actual package) across docs, examples, TypeScript extensions, CLI source, tests, and SVG assets. Also fixes `pip install emk` references to point to `agent-os-kernel[full]` since emk is a submodule, not a standalone PyPI package. Completes the fix started in PR microsoft#325 which only covered notebooks. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * docs: update framework star counts to current values Dify 65K→133K, AutoGen 42K→55K, CrewAI 28K→46K, Semantic Kernel 24K→27K, LangGraph 24K→27K, Haystack 22K→24K, Agent Framework 7.6K→8K. Added star counts for OpenAI Agents SDK (20K) and Google ADK (18K). Sorted by stars descending. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net> * fix(security): replace unregistered package names + MD5→SHA-256 - agentmesh → agentmesh-platform (5 files) - agentmesh-governance → agent-governance-toolkit - agent-os-observability → agent-os-kernel[observability] - hashlib.md5 → hashlib.sha256 (3 files) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix(scorecard): remove explicit ref in pull_request_target, add sigstore, pin CI deps Dangerous-Workflow: Remove ref: base.sha from pull_request_target workflows — default checkout IS the base branch, explicit ref triggers Scorecard false positive. Security preserved via persist-credentials: false + MSRC comments. Signed-Releases: Add sigstore/gh-action-sigstore-python to publish workflow for Python package signing alongside existing attest-build-provenance. Pinned-Dependencies: Pin pip install versions in CI workflows: - ci.yml: pytest==8.4.1, pytest-asyncio==0.26.0 - benchmarks.yml: add --no-cache-dir - security-scan.yml: pyyaml==6.0.2 - weekly-security-audit.yml: pyyaml==6.0.2 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Parsa Faraji Alamouti <165321600+parsa-faraji@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Zeel Desai <72783325+zeel2104@users.noreply.github.com> Co-authored-by: Xavier Garceau-Aranda <xavier.garceau-aranda@posteo.net>

feat: add policy evaluation heatmap to SRE dashboard (microsoft#309)

512689e

zeel2104 requested a review from imran-siddique as a code owner March 21, 2026 15:35

github-actions bot added agent-sre agent-sre package size/L Large PR (< 500 lines) labels Mar 21, 2026

github-actions bot reviewed Mar 21, 2026

View reviewed changes

imran-siddique merged commit 21b5c18 into microsoft:main Mar 21, 2026
10 checks passed

imran-siddique pushed a commit to imran-siddique/agent-governance-toolkit that referenced this pull request Mar 21, 2026

feat: add policy evaluation heatmap to SRE dashboard (microsoft#309) (m…

8ee3c81

…icrosoft#326)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: add policy evaluation heatmap to SRE dashboard (#309)#326

feat: add policy evaluation heatmap to SRE dashboard (#309)#326
imran-siddique merged 1 commit intomicrosoft:mainfrom
zeel2104:feat/policy-evaluation-heatmap

zeel2104 commented Mar 21, 2026

Uh oh!

github-actions bot commented Mar 21, 2026

Uh oh!

github-actions bot commented Mar 21, 2026

Uh oh!

github-actions bot left a comment

Uh oh!

github-actions bot commented Mar 21, 2026

Uh oh!

zeel2104 commented Mar 21, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

zeel2104 commented Mar 21, 2026

Description

Type of Change

Package(s) Affected

Checklist

Related Issues

Uh oh!

github-actions bot commented Mar 21, 2026

Uh oh!

github-actions bot commented Mar 21, 2026

🤖 AI Agent: contributor-guide

Welcome 🎉

What You Did Well ✅

Suggestions for Improvement ✨

Next Steps 🚀

Helpful Resources 📚

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

🤖 AI Agent: code-reviewer

Review Summary

🔴 CRITICAL

🟡 WARNING

💡 SUGGESTIONS

Conclusion

Uh oh!

github-actions bot commented Mar 21, 2026

🤖 AI Agent: security-scanner

Security Review of PR #309: Adding Policy Evaluation Heatmap to SRE Dashboard

1. Prompt Injection Defense Bypass

2. Policy Engine Circumvention

3. Trust Chain Weaknesses

4. Credential Exposure

5. Sandbox Escape

6. Deserialization Attacks

7. Race Conditions

8. Supply Chain

Additional Observations

Summary of Findings

Final Recommendation

Uh oh!

zeel2104 commented Mar 21, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants