The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were identified. The skill operates on local directories to assist in development workflows.
[COMMAND_EXECUTION]: The skill workflow involves executing a local script (scripts/run-checkpoints.sh) to validate generated checkpoints. This is an intended function for development validation within the skill framework.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and parses SKILL.md files from other skill directories to extract requirements.
Ingestion points: Target skill's SKILL.md, references/, and scripts/ folders.
Boundary markers: No explicit boundary markers or safety instructions for parsing untrusted markdown files were identified.
Capability inventory: Capability to write YAML files and execute local validation scripts.
Sanitization: No explicit sanitization or validation of the input skill content is mentioned.

add-checkpoints