The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute a wide range of development and security tools including ruff, mypy, pytest, eslint, and npm. While these are standard for code validation, they represent a broad execution surface within the agent context.
[DYNAMIC_EXECUTION]: The orchestration script scripts/run-all-checks.sh utilizes the eval command to execute shell strings assembled within the script. Additionally, scripts/accessibility-check.sh uses python3 -c to execute a Python snippet for JSON processing at runtime. These patterns involve the dynamic assembly and execution of code strings.
[INDIRECT_PROMPT_INJECTION]: The skill acts on untrusted project data (source code and pull request content) and has the capability to execute commands and write files, creating a potential vulnerability to instructions embedded in the files being audited.
Ingestion points: The skill reads project source code, dependency requirements, and test files in the app/, tests/, and src/ directories during automated scanning and manual review.
Boundary markers: The instructions lack delimiters or explicit directives for the agent to ignore potentially malicious instructions found within the code or data it is processing.
Capability inventory: The skill possesses Bash tool access (restricted to pytest, npm, ruff, and mypy by frontmatter but usable via the script execution) and Write tool access to create reports.
Sanitization: There is no evidence of input validation or sanitization of the file content before it is processed or interpolated into the analysis flow.

pre-merge-checklist