The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests codebase implementation details to generate documentation, creating a surface for indirect prompt injection.
Ingestion points: Reads source code, comments, and PR descriptions from the project repository as part of the documentation drafting workflow.
Boundary markers: No explicit instructions are provided to delimit ingested data or ignore instructions embedded within the codebase implementation.
Capability inventory: The skill utilizes file system analysis tools including git diff, ripgrep (rg), and ls to process project files.
Sanitization: Relies on human review of generated drafts; no automated sanitization for ingested code or metadata is defined.
[COMMAND_EXECUTION]: The skill uses shell utilities for auditing and provides command templates for incident response.
Employs git diff, ripgrep, and ls to scan for undocumented exports and documentation drift during sync audits.
Includes psql and kubectl command templates in references/runbooks-and-onboarding.md intended for manual developer use during incidents.
[EXTERNAL_DOWNLOADS]: The skill recommends an extension from the author's own repository to provide additional functionality.
Suggests adding the mermaid-diagrams skill via npx skills add oakoss/agent-skills, which targets a resource associated with the skill's author.

technical-docs