content-quality-auditor
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes content from external URLs to perform audits. This ingestion of untrusted data creates a potential surface for indirect prompt injection. The skill mitigates this by instructing the agent to treat all fetched content as 'untrusted evidence'. 1. Ingestion points: External URLs via WebFetch. 2. Boundary markers: Explicit instructions to treat text/prompts as untrusted. 3. Capability inventory: Execution of local Python scripts and file writing. 4. Sanitization: Focuses on logical separation rather than data filtering.
- [COMMAND_EXECUTION]: The skill uses shell commands like 'git rev-parse' to resolve the project root and executes local Python scripts ('rubric-score.py', 'validate-audit-artifact.py') for scoring and validation. These are standard operations within the author's ecosystem.
Audit Metadata