The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it is designed to read and audit existing codebase content, including doc comments and project files, to generate or review documentation. If the codebase being processed contains malicious instructions (e.g., hidden in code comments), they could potentially influence the agent's behavior.
Ingestion points: Uses Read, Glob, and Grep tools to ingest project files for documentation tasks.
Boundary markers: The skill does not define clear delimiters or instructions to ignore instructions embedded within the ingested code or documentation files.
Capability inventory: The agent is granted Write, Edit, Bash(go:*), Bash(git:*), Agent, and WebFetch tools, providing significant control over the filesystem and network.
Sanitization: No sanitization or validation steps are mentioned for the content retrieved from the codebase before it is processed or used to drive further agent actions.
[EXTERNAL_DOWNLOADS]: The skill references and encourages the use of various third-party tools and documentation services.
Recommends installing development utilities such as swaggo/swag, buf, and pkgsite from their respective repositories and registries.
Suggests registering projects with external AI-discovery and documentation platforms including context7.com, deepwiki.com, opendeep.wiki, and zread.ai to improve discoverability.
Includes documentation templates that show users how to provide binary downloads via GitHub Releases using curl commands.

backend-go-documentation