Skip to content

feat: add concept-diagrams skill (salvage of #11045, v1k22)#11363

Merged
teknium1 merged 4 commits into
mainfrom
hermes/hermes-771b5dc3
Apr 17, 2026
Merged

feat: add concept-diagrams skill (salvage of #11045, v1k22)#11363
teknium1 merged 4 commits into
mainfrom
hermes/hermes-771b5dc3

Conversation

@teknium1

Copy link
Copy Markdown
Contributor

Salvage of #11045 from @v1k22. Introduces a new optional skill for generating flat, minimal SVG diagrams of physics / chemistry / math / engineering concepts, physical objects, and system flows — complementing the existing architecture-diagram skill (dark, software-infra focused).

What this adds

optional-skills/creative/concept-diagrams/:

  • SKILL.md — design system + diagram types + output guidance
  • templates/template.html — host page with full CSS (9 color ramps, light/dark auto)
  • references/physical-shape-cookbook.md — shape mapping for physical diagrams
  • references/infrastructure-patterns.md — hub-spoke + power/water/transport primitives
  • references/dashboard-patterns.md — UI mockup building blocks
  • examples/*.md — 15 complete example diagrams (aircraft, SN2 reaction, wind turbine, smart city, CPU microarchitecture, etc.)

Install: hermes skills install official/creative/concept-diagrams

Changes on top of the original PR

Preserves @v1k22's commit and authorship. Follow-up commit tightens the skill for shipping:

  • Rename architecture-visualization-svg-diagramsconcept-diagrams. Differentiates clearly from the existing architecture-diagram skill (dark-themed Cocoon-style) and captures the broader educational scope v1k22 pitched in the thread (physics, chemistry, math, engineering).
  • Scope the trigger description. Removed the 'always use this skill' language and the long phrase-capture list to stop colliding with architecture-diagram, excalidraw, generative-widgets, and manim-video. Description now points users at the intended use cases.
  • Default to standalone HTML output. The preview server is opt-in and no longer part of the default workflow. Most users want a single .html file they can open directly.
  • Fix server security/collision issues (when the server is used):
    • Bind to 127.0.0.1 instead of 0.0.0.0 — the original would have exposed user diagrams on every network interface.
    • Let the OS pick a free ephemeral port instead of hard-coding 22223 (collision prone across users/agents).
    • Document how to stop the server.
  • Shrink SKILL.md from 1540 to 353 lines by extracting reusable material into linked files (templates/template.html, three references/*.md files). Skills are always loaded fully, so this meaningfully reduces per-session token cost while keeping all 15 examples and design rules intact.
  • Add dhandhalyabhavik@gmail.com → v1k22 to scripts/release.py's AUTHOR_MAP.

Test plan

  • tests/tools/test_skills* and related — 228 passed, 1 skipped
  • E2E install via hermes skills install official/creative/concept-diagrams — all 20 files (SKILL.md + template + 3 references + 15 examples) land correctly under skills/creative/concept-diagrams/
  • Rendered a sanity-check diagram through the template and verified well-formed HTML output

Closes #11045

Credit to @v1k22 for the original skill, design system, and all 15 example diagrams.

v1k22 and others added 4 commits April 16, 2026 20:16
…ategory

- SKILL.md with full SVG design system (color palette, typography, spacing, dark mode)
- 15 example diagrams covering flowcharts, physical structures, chemistry, charts, floor plans, and more
- Supports 8 diagram types: flowchart, structural, API map, microservice, data flow, physical, infrastructure, UI mockups
- Auto-hosts diagrams on 0.0.0.0:22223 as interactive web pages
Salvage of PR #11045 (original by v1k22). Changes on top of the
original commit:

- Rename 'architecture-visualization-svg-diagrams' -> 'concept-diagrams'
  to differentiate from the existing architecture-diagram skill.
  architecture-diagram stays as the dark-themed Cocoon-style option for
  software/infra; concept-diagrams covers physics, chemistry, math,
  engineering, physical objects, and educational visuals.
- Trigger description scoped to actual use cases; removed the 'always
  use this skill' language and long phrase-capture list to stop
  colliding with architecture-diagram, excalidraw, generative-widgets,
  manim-video.
- Default output is now a standalone self-contained HTML file (works
  offline, no server). The preview server is opt-in and no longer part
  of the default workflow.
- When the server IS used: bind to 127.0.0.1 instead of 0.0.0.0 (was a
  LAN exposure hazard on shared networks) and let the OS pick a free
  ephemeral port instead of hard-coding 22223 (collision prone).
- Shrink SKILL.md from 1540 to 353 lines by extracting reusable
  material into linked files:
    - templates/template.html (host page with full CSS design system)
    - references/physical-shape-cookbook.md
    - references/infrastructure-patterns.md
    - references/dashboard-patterns.md
  All 15 examples kept intact.
- Add dhandhalyabhavik@gmail.com -> v1k22 to AUTHOR_MAP.

Preserves v1k22's authorship on the underlying commit.
…xplicit

Both skills generate SVG system diagrams, but for very different subjects
and aesthetics. The old descriptions didn't make the split clear, so an
agent loading either one couldn't confidently pick.

Changes:

- Rewrote both frontmatter descriptions to state the scope up front plus
  an explicit 'for X, use the other skill instead' pointer.
- Added a symmetric 'When to use this skill vs <other>' decision table
  to the top of each SKILL.md body, so the guidance is visible whether
  the agent is reading frontmatter or full content.
- Added architecture-diagram <-> concept-diagrams to each other's
  related_skills metadata.

Rule of thumb baked into both skills:
  software/cloud infra -> architecture-diagram
  physical / scientific / educational -> concept-diagrams
Previous pass assumed both skills would always be loaded together, so
each description pointed at the other ('use concept-diagrams instead').
That breaks when only one skill is active — the agent reads 'use the
other skill' and there is no other skill.

Now each skill's description and scope section is fully self-contained:

- States what it's best suited for
- Lists subjects where a more specialized skill (if available) would be
  a better fit, naming them only as 'consider X if available'
- Explicitly offers itself as a general SVG diagram fallback when no
  more specialized skill exists

An agent loading either skill alone gets unambiguous guidance; an
agent with both loaded still gets useful routing via the 'consider X
if available' hints and the related_skills metadata.
@teknium1 teknium1 merged commit fdefd98 into main Apr 17, 2026
4 of 5 checks passed
@teknium1 teknium1 deleted the hermes/hermes-771b5dc3 branch April 17, 2026 03:39
teknium1 added a commit that referenced this pull request Apr 18, 2026
Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (#11363) — optional-skills-catalog entry
- /gquota (#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
teknium1 added a commit that referenced this pull request Apr 18, 2026
Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (#11363) — optional-skills-catalog entry
- /gquota (#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
ulasbilgen pushed a commit to ulasbilgen/hermes-adhd-agent that referenced this pull request May 1, 2026
)

Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (NousResearch#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (NousResearch#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (NousResearch#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (NousResearch#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (NousResearch#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (NousResearch#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (NousResearch#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (NousResearch#11363) — optional-skills-catalog entry
- /gquota (NousResearch#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
aj-nt pushed a commit to aj-nt/hermes-agent that referenced this pull request May 1, 2026
)

Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (NousResearch#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (NousResearch#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (NousResearch#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (NousResearch#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (NousResearch#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (NousResearch#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (NousResearch#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (NousResearch#11363) — optional-skills-catalog entry
- /gquota (NousResearch#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
02356abc pushed a commit to 02356abc/hermes-agent that referenced this pull request May 14, 2026
)

Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (NousResearch#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (NousResearch#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (NousResearch#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (NousResearch#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (NousResearch#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (NousResearch#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (NousResearch#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (NousResearch#11363) — optional-skills-catalog entry
- /gquota (NousResearch#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
gweeteve pushed a commit to gweeteve/hermes-agent that referenced this pull request Jun 2, 2026
)

Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (NousResearch#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (NousResearch#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (NousResearch#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (NousResearch#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (NousResearch#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (NousResearch#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (NousResearch#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (NousResearch#11363) — optional-skills-catalog entry
- /gquota (NousResearch#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
Egavasyug pushed a commit to Egavasyug/hermes-agent that referenced this pull request Jun 10, 2026
)

Fills documentation gaps that accumulated as features merged ahead of their
docs updates. All additions are verified against code and the originating PRs.

Providers:
- Ollama Cloud (NousResearch#10782) — new provider section, env vars, quickstart/fallback rows
- xAI Grok Responses API + TTS (NousResearch#10783) — provider note, TTS table + config
- Google Gemini CLI OAuth (NousResearch#11270) — quickstart/fallback/cli-commands entries
- NVIDIA NIM (NousResearch#11774) — NVIDIA_API_KEY / NVIDIA_BASE_URL in env-vars reference
- HERMES_INFERENCE_PROVIDER enum updated

Messaging:
- DISCORD_ALLOWED_ROLES (NousResearch#11608) — env-vars, discord.md access control section
- DingTalk QR device-flow (NousResearch#11574) — wizard path in Option A + openClaw disclosure
- Feishu document comment intelligent reply (NousResearch#11898) — full section + 3-tier access control + CLI

Skills / commands:
- concept-diagrams skill (NousResearch#11363) — optional-skills-catalog entry
- /gquota (NousResearch#11270) — slash-commands reference

Build: docusaurus build passes, ascii-guard lint 0 errors.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants