The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes content from documentation fragments and git diffs which could potentially contain malicious instructions.
Ingestion points: Markdown files located in the docs/context/ directory and git diff output.
Boundary markers: Absent; the instructions do not implement specific delimiters or 'ignore' directives for the content of the documentation fragments.
Capability inventory: The skill utilizes version control commands (git), executes local scripts, and performs file writes to documentation files.
Sanitization: Absent; content from the documentation fragments is integrated into the session context or processed into indexes without specific security filtering.
[COMMAND_EXECUTION]: The skill uses version control commands to manage session state and context.
Evidence: SKILL.md directs the agent to run git log --oneline -15 and git status to establish orientation in a new session.
[DYNAMIC_EXECUTION]: The skill executes local scripts provided in its distribution to maintain its documentation mapping.
Evidence: The skill triggers bash scripts/drift.sh to identify documentation drift and python3 scripts/build-map.py to regenerate MAP.md and README.md.

tools-registry-context