game-changing-features
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local codebase and user-provided inputs without sanitization. Ingestion points: The agent is instructed to research the codebase and process user-supplied product descriptions and constraints. Boundary markers: Absent; there are no instructions to ignore embedded directives in the files read. Capability inventory: The skill has file-read access to the codebase and file-write access to the project directory. Sanitization: Absent; no validation or escaping of external content is mentioned.
- [PROMPT_INJECTION]: The skill instructs the agent to suppress all standard chat output, which can be used to conceal actions from the user. Evidence: The directive 'No Chat Output: ALL responses go to .claude/docs/ai//10x/session-N.md' redirects communication to a file.
- [COMMAND_EXECUTION]: The skill uses a user-provided variable to construct a file path for output, creating a risk of path traversal. Evidence: The output path includes the '' placeholder provided by the user during setup.
Audit Metadata