The Agent Skills Directory

[SAFE]: The skill implements a logical triage workflow using a state machine (needs-triage, needs-info, ready-for-agent, etc.). It manages issue labels and provides templates for communication with contributors.
[SAFE]: The skill uses a .out-of-scope/ directory to store project-specific decisions as markdown files, which is a standard documentation practice for open-source maintenance.
[SAFE]: The AGENT-BRIEF.md documentation provides best practices for creating clear, behavioral specifications for AI agents without using dangerous command execution or sensitive file access.
[SAFE]: The configuration in agents/openai.yaml correctly restricts implicit invocation, ensuring the agent only acts when explicitly requested by a maintainer.

triage