The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/update_genome.py utilizes subprocess.run to execute git log commands. This is used to synthesize a summary of recent repository changes for the recent_changes section of the genome. The implementation uses a structured list of arguments rather than a raw shell string, which is a safe practice.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates an attack surface where instructions embedded in project documentation could influence the agent during the analysis phase.
Ingestion points: The agent is instructed to read the first 3000 characters of all discovered .md files in the repository to evaluate their importance and purpose.
Boundary markers: There are no explicit delimiters or system instructions provided to the agent to disregard instructions found within the documentation files during analysis.
Capability inventory: The skill allows the agent to execute Python scripts (which in turn run git commands) and write to critical project files like PROJECT-GENOME.yaml and CLAUDE.md.
Sanitization: The skill does not implement content sanitization or instruction-filtering for the markdown data it ingests before presenting it to the agent for analysis.

project-genome