design-optimize
Pass
Audited by Gen Agent Trust Hub on Mar 23, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill reads project configuration files and source code to identify optimization opportunities. This data is untrusted and could contain embedded instructions intended to manipulate the agent's behavior.
- Ingestion points: Reads project manifests (package.json, pubspec.yaml, build.gradle) and source files (CSS, JS, TS, Flutter) in SKILL.md.
- Boundary markers: None; the skill does not use explicit delimiters or instructions to ignore embedded prompts when reading codebase files.
- Capability inventory: The skill can modify source code (Phase 8), run build commands (Phase 9.1), and execute project tests (Phase 9.2) in SKILL.md.
- Sanitization: No evidence of sanitization or validation of the content read from files before it is processed by the agent.
- [COMMAND_EXECUTION]: Functional Command Execution. The skill performs build and test verification (Phase 9.1 and 9.2 in SKILL.md), which involves executing shell commands to verify optimizations. While consistent with its stated purpose, this provides an execution surface for potential injection attacks.
Audit Metadata