The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to interact with the local environment, including git diff and git status for change identification. It also dynamically detects and executes the project's native test runner (e.g., pytest, npm test, go test) by inspecting configuration files like package.json, pyproject.toml, or CI workflow definitions.
[DYNAMIC_EXECUTION]: The core workflow involves the agent writing new executable code (test files) and then running that code against the local codebase. This process is managed by a provided script, scripts/prove-catch.sh, which executes arbitrary commands provided by the agent to verify test failures against 'mutant' versions of the code.
[INDIRECT_PROMPT_INJECTION]: This skill presents an attack surface where a malicious repository could influence the agent.
Ingestion points: The skill reads source code and existing tests from the repository (SKILL.md Step 1 and 2).
Boundary markers: No explicit markers are used to separate repository content from agent instructions during analysis.
Capability inventory: The skill possesses file-write, file-read, and command execution capabilities across all its scripts.
Sanitization: The skill does not perform sanitization of the code it reads before using it to generate new test logic, potentially allowing malicious code comments or patterns to influence the generated output or the agent's reasoning.

review-jit-test-gen