test-driven-development
Pass
Audited by Gen Agent Trust Hub on Mar 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run
npm testto verify code behavior during the TDD cycle. This involves executing local scripts and tests within the environment. - [PROMPT_INJECTION]: The skill uses extremely strict behavioral instructions ('The Iron Law', 'Delete means delete', 'No exceptions') to override typical agent behavior when code is written without tests. While focused on TDD methodology, these patterns resemble behavioral override techniques used to force an agent to ignore its standard instructions or previous work.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. 1. Ingestion points: The agent ingests and processes user-provided source code, test files, and bug reports as part of the TDD cycle. 2. Boundary markers: The process is structured into logical phases (Red-Green-Refactor), but lacks formal delimiters to prevent the agent from following instructions embedded in the code it analyzes. 3. Capability inventory: The agent can execute shell commands (
npm test) and is instructed to modify/delete code blocks based on its analysis. 4. Sanitization: There is no explicit sanitization or validation of the code/test content before it is executed or processed.
Audit Metadata