The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script (scripts/check_profile.py) to validate the structure and activation status of the user's 'person-dna.md' profile. The script performs local regex-based validation and does not exhibit any malicious behavior such as network communication or unauthorized file modification.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided content via 'person-dna.md' and selected topics.
Ingestion points: Data enters through the person-dna.md file and user topic inputs as described in SKILL.md.
Boundary markers: The skill establishes clear boundaries in references/voice-and-evidence.md, instructing the agent to only read explicitly specified files and to maintain a distinction between user facts and instructional frameworks.
Capability inventory: The skill is limited to text generation and lacks dangerous capabilities such as network access or file-system write permissions.
Sanitization: The scripts/check_profile.py utility validates the structural integrity and activation status of the profile before the agent processes the content.

yetong-copy