iterative-development
Pass
Audited by Gen Agent Trust Hub on Apr 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides Bash scripts intended for execution via 'Stop', 'PreToolUse', and 'SessionStart' hooks. These scripts execute various local development tools including 'npm test', 'npm run lint', 'npx tsc', 'pytest', 'ruff', and 'mypy'.
- [PROMPT_INJECTION]: The iteration loop captures the stdout and stderr of test and lint commands and feeds it directly back to the agent to trigger automated fixes. 1. Ingestion points: Command output variables (TEST_OUTPUT, LINT_OUTPUT, TYPE_OUTPUT) in 'scripts/tdd-loop-check.sh'. 2. Boundary markers: Absent; command output is sent to stderr without delimiters or instructions to ignore embedded instructions. 3. Capability inventory: The agent has capabilities for file system modification and command execution within the project environment. 4. Sanitization: Absent; output is truncated using 'tail' but is not escaped or validated, allowing potential malicious output from repository files to influence agent behavior.
Audit Metadata