add-feature
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates user-provided feature descriptions via the $ARGUMENTS parameter directly into its operational logic. Without proper boundary markers or sanitization, a malicious user could provide a description that acts as a prompt injection, potentially tricking the agent into performing unauthorized actions such as exfiltrating data or introducing backdoors during the code generation process.
- Ingestion points: User input provided via the $ARGUMENTS placeholder in SKILL.md.
- Boundary markers: None present; user input is not delimited from system instructions.
- Capability inventory: The skill can read source files, create/modify files in the src/ directory, and execute shell commands (e.g., npm run build).
- Sanitization: None; input is interpolated directly.
- [COMMAND_EXECUTION]: The skill performs shell command execution to verify the integrity of the feature implementation.
- Evidence: Execution of npm run build to confirm no errors after code modification in SKILL.md.
Audit Metadata