The Agent Skills Directory

[COMMAND_EXECUTION]: The reference/bug-fix.md file explicitly instructs the agent to execute verification commands after applying fixes, specifically referencing npm test, pytest, go test ./..., eslint, ruff, and golangci-lint.
[EXTERNAL_DOWNLOADS]: In reference/implementation.md, the agent is prompted to prepare the environment by installing dependency packages, which involves fetching code from external repositories like NPM or PyPI.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its processing of untrusted user data combined with its command execution capabilities.
Ingestion points: The agent ingests user-provided bug reports, logs, and code snippets within reference/bug-fix.md, reference/implementation.md, and reference/review.md.
Boundary markers: Absent. The skill does not provide delimiters or specific instructions for the agent to distinguish between its own system instructions and potentially malicious instructions embedded in the user-provided data.
Capability inventory: The agent is granted the ability to modify local files and execute system commands including package managers, linters, and test runners.
Sanitization: Absent. There are no mechanisms defined to validate or sanitize user input before it is utilized in file writing or subprocess execution.

dev-workflow