kw-optimize-agents-project

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local shell commands to analyze project documentation. Evidence: The workflow in SKILL.md instructs the agent to run wc -l, wc -w, and several grep operations on AGENTS.md and CLAUDE.md to baseline heading counts and identify GSD markers.- [INDIRECT_PROMPT_INJECTION]: The skill reads and refactors repository documentation files, creating a surface for indirect instructions. Ingestion points: The skill reads the target instruction file (AGENTS.md or CLAUDE.md) in Step 1 of SKILL.md. Boundary markers: No specific delimiters or safety instructions are used to distinguish ingested content from agent commands. Capability inventory: The skill has local file system read/write access and the ability to execute basic text analysis commands. Sanitization: The skill relocates blocks of text verbatim without validating or escaping the content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 09:15 AM
Security Audit — agent-trust-hub — kw-optimize-agents-project