prd
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data that could influence the agent's behavior during the PRD generation process.
- Ingestion points: User input via $ARGUMENTS and codebase content accessed via the 'Explore codebase' instruction in SKILL.md.
- Boundary markers: Absent. The skill does not use specific delimiters or instructions to ignore embedded commands within the ingested data.
- Capability inventory: The skill performs codebase exploration (file reading) and saves the generated PRD to the filesystem (file writing) as seen in Step 2 and Step 5.
- Sanitization: The skill applies a 'slug' derivation logic to user input (stripping verbs, lowercasing, removing fillers, and limiting word count) before using it in a filename, which provides protection against path traversal.
- [COMMAND_EXECUTION]: The skill requires access to the local environment to read source code files and write the resulting documentation to .specs/prds/.md. These actions are performed using the agent's default tools and are consistent with the skill's stated purpose of PRD generation and codebase mapping.
Audit Metadata