context-collectorlisted

<SUBAGENT-STOP> If you were dispatched as a subagent to execute a specific task, skip this skill entirely. This skill is for orchestrators and interactive session-level routing only. Proceed directly with your assigned task. </SUBAGENT-STOP> # Context Collector ## Purpose Collects and maintains a **unified context document** for a job. The context document is a curated, focused summary that all sub-agents can reference — it eliminates redundant research and ensures consistent understanding of the task across the entire job pipeline. **Key responsibilities:** 1. Gather relevant local documentation (`docs/`, `jobs/`, `rules/core/`) 2. Identify libraries, patterns, and references used in the task area 3. Fetch external documentation (library docs, API references, best practices) via web 4. Extract and summarize only what is relevant to the specific task 5. Assemble everything into a single `context.md` document 6. Keep the context document up-to-date as the job evolves (version-tracked updates) **Input:** Task description + job name + project path (from orchestrator or user) **Output:** `context.md` document in `.metaproject/jobs/<job-name>/ai/` and `.metaproject/jobs/<job-name>/man/`, documented via `job-documenter` ## When to Use - Called by `job-orchestrator` as an early plan step (after analysis, before implementation) - Called directly by user when they need a research/context document for a task - Called again during a job when additional context is needed (e.g., ne