The Agent Skills Directory

[PROMPT_INJECTION]: The skill defines capabilities for processing and testing external documentation snippets and API specifications, which introduces a surface for indirect prompt injection.
Ingestion points: API specifications, code comments, and documentation snippets (SKILL.md, openapi-spec-generation.md).
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the ingested content.
Capability inventory: Includes capabilities for automated testing of code snippets and execution of curl commands found in documentation (SKILL.md).
Sanitization: There are no documented procedures for validating or escaping content from external snippets before processing.
[EXTERNAL_DOWNLOADS]: The reference documentation includes instructions for installing various open-source developer tools and libraries from official package registries (e.g., npm install husky, pip install commitizen). These references target well-known technology services and are part of standard documentation and release management workflows.
[COMMAND_EXECUTION]: Multiple examples of CLI commands are provided for project initialization, documentation generation (e.g., git cliff, openapi-generator-cli), and linting. These are presented as instructional examples for the user or for an agent to suggest in a development context.
[SAFE]: The skill follows security best practices in its templates, such as demonstrating the use of repository secrets in GitHub Actions workflows rather than hardcoding credentials.

api-documenter