domains / families / agent-tooling / instructree

instructree

authentic

Map, explain, and lint repository-scoped coding-agent instructions before changing code.

When to use

• Use before changing code in a repository with `AGENTS.md`, `CLAUDE.md`, Copilot instructions, agent skills, custom agents, Cursor rules, or Windsurf rules.
• Use when the user asks which instructions may apply to one target file.
• Use when auditing recursive Copilot CLI `@path` imports or exporting instruction diagnostics to CI.

When not to use

• Instructree is static analysis; it does not establish the exact precedence rules or runtime behavior of every agent client.
• Client discovery and import behavior can change, so `explain` reports what may apply rather than predicting what a model will follow.
• The audit does not prove that instruction content is correct, safe, or effective.
• Do not edit instruction files unless the user asked for changes.

Skill page

Source