AgentSkillsCN

archivist-retriever

主动搜索并应用 `./reference-material/docs` 中的本地文档。

SKILL.md
--- frontmatter
name: archivist-retriever
description: Proactively searches and applies local documentation from `./reference-material/docs`.

Archivist Retriever

Prioritize local documentation over internal knowledge for every technical query.

Triggers

  • Any task involving libraries, tools, or technical concepts (e.g., "How do I use PandaCSS?", "Implement a TanStack Router").
  • Direct questions about how something works within the project's tech stack.

Core Mandates

  • Library First: You MUST check ./reference-material/docs for information before relying on your internal knowledge or performing external searches.
  • Exhaustive Search: Use glob and search_file_content to find relevant information in the local documentation library.
  • Contextual Application: Apply the found documentation precisely to the current task, adhering to the documented patterns and versions.

Workflow: Retrieval & Activation

  1. Identify Need: Recognize when a task involves a technology or concept likely documented in ./reference-material/docs.
  2. Search Library:
    • Use glob to find relevant files in ./reference-material/docs/.
    • Use search_file_content to find specific keywords or symbols within those files.
  3. Analyze & Inform:
    • Read the relevant documentation files.
    • If information is found, explicitly state: "Found relevant documentation in ./reference-material/docs/. Applying these patterns..."
  4. Apply Knowledge: Proceed with the task using the retrieved documentation as the primary source of truth.
  5. Fallback: If the information is not found or is insufficient, explicitly state: "I could not find sufficient information in the local library. Activating archivist-researcher for further investigation."