Skip to content

AGENTS.md

Latest commit

 

History

History
35 lines (26 loc) · 1.43 KB

AGENTS.md

File metadata and controls

35 lines (26 loc) · 1.43 KB

AGENTS.md — Bootstrap

See AGENT_INSTRUCTIONS.md for complete instructions.

Prime Protocol

On every session start, after compaction, or after context clear:

  1. Run the repo's prime command: make agent-prime or ./scripts/agent-prime.sh
  2. If no prime command exists, read AGENT_INSTRUCTIONS.md and identify test/lint/build commands before editing code.
  3. Re-prime after any compaction or context reset.

Skill Router

After prime, classify your task and load the relevant skill(s):

  • Architecture, testing, or security decisions → engineering-standards
  • Adding or evaluating a dependency → dependency-adding
  • Security-sensitive changes (auth, crypto, secrets) → security-review
  • CI/CD pipeline failure → ci-debugging
  • Writing a feature spec → feature-spec
  • Air-gapped or disconnected deployment → disconnected-environments
  • Architecture decision to record → adr-writing
  • Debugging or recurring failure → root-cause-analysis
  • MCP server integration → mcp-integration
  • Whole-repo code health assessment → code-health-review
  • Writing or editing AGENTS.md / CLAUDE.md → context-file-authoring
  • Correction that applies to all projects → upstream-contribution

Ask Before Proceeding

  • Requirements are ambiguous.
  • The change affects public APIs or security posture.
  • A new production dependency is needed.
  • The change requires broad refactors or file moves.