optimise-claude

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute the 'bun skills' command to synchronize optimized skill files. This is a standard part of the documented workflow for this skill as described in the README.md and SKILL.md files.\n- [PROMPT_INJECTION]: The skill processes the content of other skill files and 'AGENTS.md', creating a surface for indirect prompt injection.\n
  • Ingestion points: Reads content from '.agents/skills/*/SKILL.md' and 'AGENTS.md' during the inventory, reduction, and deduplication phases.\n
  • Boundary markers: No delimiters or 'ignore' instructions are specified when processing external file contents.\n
  • Capability inventory: The skill has the ability to write new skill files to '.agents/skills/' and execute local shell commands (bun skills).\n
  • Sanitization: No content validation or escaping is performed on the data read from external skill files before it is processed or used to create new skills.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 10:15 AM