The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill's core logic is designed to increase agent reliability by requiring bounded contracts and authoritative evidence for task completion while maintaining human-in-the-loop oversight.- [EXTERNAL_DOWNLOADS]: The README documentation provides installation commands using standard package runners (npx, bunx) to add the skill from the author's repository (dennexequiel/skills). These are recognized as legitimate vendor resources and standard platform distribution methods.- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests untrusted data from the local environment and command outputs to perform its verification duties. However, the design includes significant mitigations.
Ingestion points: The skill reads file contents and captures command exit statuses and outputs across the mission lifecycle.
Boundary markers: The skill employs a formal 'Mission Brief' structure and terminal state reporting to separate external data from mission instructions.
Capability inventory: The skill is authorized to perform routine implementation, diagnostic, and verification cycles via terminal commands and file edits within the agreed mission scope.
Sanitization: The skill relies on authoritative verification evidence and human-in-the-loop checkpoints for consequential decisions rather than explicit input filtering.

ace