The Agent Skills Directory

[SAFE]: The skill implements a structured documentation process using specific tags like ARCH, SEC, and IRREV to ensure architectural, security, and irreversible changes are logged and reviewed.
[COMMAND_EXECUTION]: The provided scripts in references/canary-signals.md perform repository analysis using standard command-line tools such as git, jq, and grep. These are intended for automated detection of dependency and interface changes within a CI/CD environment.
[PROMPT_INJECTION]: The skill defines how to process markers in code. While the agent reads source code to extract these markers, the instructions focus on logging and classification, which is standard for development-oriented agents and does not bypass safety filters.

documenting-decisions