sanitize-artifacts
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The primary goal of the skill is to improve the professional presentation of deliverables by stripping out production-related artifacts and conversation history.
- [PROMPT_INJECTION]: The instructions include a concealment pattern by directing the agent to suppress explanations about the changes it makes (e.g., 'Do not preface the artifact with process commentary'). This reduces transparency for the user regarding the agent's modifications.
- [PROMPT_INJECTION]: The skill acts as a processing engine for session artifacts, creating a potential surface for indirect prompt injection from untrusted content. Ingestion points: Text artifacts produced during the session as described in SKILL.md. Boundary markers: The skill does not implement specific delimiters or instructions to treat ingested artifacts as non-executable data. Capability inventory: The agent is granted the capability to analyze, modify, and output revised text based on the ingested content. Sanitization: The logic focuses on stylistic sanitization (removing 'residue') but lacks security-oriented sanitization for malicious instructions hidden in the data.
Audit Metadata