feat(agents): add the-user coordinator agent#2
Conversation
Introduces a new lightweight coordinator agent that acts as an intelligent intermediary between users and specialized AI agents. Key features: - Uses cheaper models (haiku/flash/nano) for cost-effective coordination - Understands user intent and routes to appropriate agents - Relays questions from agents back to users for decision-making - Includes comprehensive system prompt with decision trees and examples - Supports session continuation for multi-turn agent interactions https://claude.ai/code/session_01VK8KpJng69ry31LhwuFmzp
…n workflow Enhances the-user coordinator agent with Planning & Execution workflow: - Prometheus integration for strategic planning of complex features - Atlas for structured execution with verification and parallel tasks - Hephaestus for autonomous deep work and complex exploration - Decision tree updated to route big features through planning workflow - Added examples for planning workflow and executor selection - Updated boundaries section with clear agent responsibilities https://claude.ai/code/session_01VK8KpJng69ry31LhwuFmzp
…ork only) Atlas is reserved for the /start-work command. Updated the-user agent to: - Use only Hephaestus for plan execution - Remove Atlas from decision tree and examples - Add note clarifying Atlas is /start-work only https://claude.ai/code/session_01VK8KpJng69ry31LhwuFmzp
📝 WalkthroughWalkthroughA new "the-user" coordinator agent is introduced, orchestrating user requests by delegating to specialized agents and aggregating results. The implementation spans type definitions, schema validation, agent utilities, model configuration, and a primary agent file with dynamic prompt generation and delegation logic. Changes
Sequence DiagramsequenceDiagram
participant User as User
participant Coordinator as The User Coordinator
participant Agents as Specialized Agents<br/>(Prometheus, Hephaestus, etc.)
User->>Coordinator: Send request with category/intent
Note over Coordinator: Parse request<br/>Determine category
Coordinator->>Agents: Delegate to specialized agent(s)<br/>based on category & skills
Agents-->>Coordinator: Agent response
Coordinator->>Coordinator: Aggregate & format results
Coordinator-->>User: Return orchestrated response
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Poem
🚥 Pre-merge checks | ✅ 1 | ❌ 2❌ Failed checks (2 warnings)
✅ Passed checks (1 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Update all fallback chains to match current model-requirements.ts: - Librarian: now minimax-m2.5-free -> gemini-flash -> big-pickle (free-tier first) - Explore: add minimax-m2.5-free as #2 after grok-code-fast-1 - Multimodal Looker: reorder to kimi-first (k2p5 -> kimi-free -> flash -> gpt-5.2) - Atlas: remove gemini-3-pro, keep kimi k2.5 -> sonnet -> gpt-5.2 - GLM 4.7 -> GLM 5 everywhere - Add venice provider for grok, opencode provider for glm-5 Add design philosophy section explaining the intelligence hierarchy: premium models for core agents, free-tier for utility agents, balanced for orchestrators. Document why utility agents intentionally use cheap models and why Kimi K2.5 appears as primary for multiple agents.
- Create gpt.ts with XML-tagged, principle-driven prompt (Codex plan mode style) - Add getPrometheusPrompt() routing: GPT models → GPT prompt, others → default - Promote gpt-5.2 (high) to #2 in prometheus fallback chain - Follow Atlas GPT variant pattern (isGptModel detection)
Restructure from 13 scattered XML blocks to 8 dense blocks with 9 named sub-anchors, following OpenAI GPT-5.4 prompting guidance and Oracle-reviewed context preservation strategy. Key changes: - Merge think_first + intent_gate + autonomy into unified <intent> with domain_guess classification and <ask_gate> sub-anchor - Add <execution_loop> as central workflow: EXPLORE -> PLAN -> ROUTE -> EXECUTE_OR_SUPERVISE -> VERIFY -> RETRY -> DONE - Add mandatory manual QA in <verification_loop> (conditional on runnable behavior) - Move <constraints> to position #2 for GPT-5.4 attention pattern - Add <completeness_contract> as explicit loop exit gate - Add <output_contract> and <verbosity_controls> per GPT-5.4 guidance - Add domain_guess (provisional) in intent, finalized in ROUTE after exploration -- visual domain always routes to visual-engineering - Preserve all named sub-anchors: ask_gate, tool_persistence, parallel_tools, tool_method, dependency_checks, verification_loop, failure_recovery, completeness_contract - Add skill loading emphasis at intent/route/delegation layers - Rename EXECUTE to EXECUTE_OR_SUPERVISE to preserve orchestrator identity with non-execution exits (answer/ask/challenge)
Introduces a new lightweight coordinator agent that acts as an intelligent
intermediary between users and specialized AI agents. Key features:
https://claude.ai/code/session_01VK8KpJng69ry31LhwuFmzp
Summary by CodeRabbit
Release Notes