The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute shell commands for git operations (git diff, git ls-files) and to run package manager commands (pnpm, npm, yarn, pip, go). It also executes a dynamically determined RUN_COMMAND to run the generated test suites, which could lead to arbitrary command execution if the project's configuration (e.g., package.json scripts) is maliciously altered.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and installation of development dependencies such as Vitest, Playwright, Cypress, and Pytest from public registries. While these are well-known and standard tools, the process involves downloading external code that is then executed in the local environment.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) due to its core functionality of reading and interpreting untrusted local files.
Ingestion points: The agent reads uncommitted source files, specification files located in docs/specs/, and design.md files to generate test logic.
Boundary markers: The instructions lack explicit boundary markers or directives to ignore instructions embedded within the source code or documentation being analyzed.
Capability inventory: The skill possesses broad capabilities, including the ability to write to the file system, execute shell commands, and orchestrate subagents.
Sanitization: There is no evidence of sanitization or filtering applied to the content of the files under test before they are ingested into the agent's context.

test