The Agent Skills Directory

[COMMAND_EXECUTION]: The test suite in tests/test_evaluation_record.py uses subprocess.run() to invoke the skill's own utility script scripts/evaluation_record.py. This execution is restricted to the local Python interpreter and known script paths for testing purposes.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze conversation history and external material, which are untrusted sources. It addresses the potential for indirect prompt injection by explicitly instructing the agent to treat this data as evidence rather than instructions and by providing a sanitization framework in references/retention-and-migration.md to remove sensitive information.
[DATA_EXFILTRATION]: While the skill manages and records history, it uses dedicated integrity and path-validation logic in scripts/evaluation_record.py and scripts/history_reuse.py to ensure all file operations are confined to the intended local history directory and verified against expected hashes.

evolve-skills