The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill gathers deployment context and environment details from the user to generate CI/CD configuration files, which are then saved to the repository. This represents a vulnerability surface where malicious user input could potentially inject instructions into the generated configurations.
Ingestion points: Phase 1 detection questions in SKILL.md regarding platforms, CI/CD systems, and infrastructure.
Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard potentially malicious instructions embedded in the user-supplied data.
Capability inventory: The skill possesses the capability to write files to the repository, specifically for saving deployment configurations in Phase 5.
Sanitization: The skill does not include steps for sanitizing or validating user-provided environment details before they are interpolated into configuration templates.
[EXTERNAL_DOWNLOADS]: The skill's configuration examples include references to official utility actions from a well-known service for repository checkout and runtime setup.
Evidence: YAML templates in Phase 3 of SKILL.md reference actions/checkout@v4 and actions/setup-node@v4 from the official GitHub Actions organization.

deployment