The Agent Skills Directory

[SAFE]: The skill functions as a specialized assistant for writing technical documentation. It provides detailed frameworks for various document types including README.md, OpenAPI 3.1 specifications, tutorials, Architecture Decision Records (ADRs), and changelogs.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data, such as source code and project descriptions, to generate documentation.
Ingestion points: User-provided code snippets or repository URLs provided as context for documentation generation in SKILL.md.
Boundary markers: None defined to isolate user-provided code from the documentation instructions.
Capability inventory: The skill is limited to text generation (Markdown and YAML). It does not request or use tools for network access, file system modifications, or shell command execution.
Sanitization: No specific input sanitization is performed. The risk is considered minimal due to the lack of actionable capabilities that an injection could exploit.
[CREDENTIALS_UNSAFE]: The documentation templates (templates/output-template.md) and examples (examples/example-output.md) include placeholders for environment variables such as API_URL and GENPASS_WORDLIST. These are used for illustrative purposes only and do not contain real secrets or encourage insecure handling of sensitive data.
[COMMAND_EXECUTION]: The skill documentation includes bash examples for installation (e.g., npm install -g genpass) and usage. These commands are provided as static text to be included in generated documentation for end-users and are not executed by the AI agent as part of the skill's operational logic.

tech-writer