The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it ingests data from untrusted local files such as source code, API specifications, and git history.
Ingestion points: Processes project files in SKILL.md (Step 2) and various stack-specific reference files (e.g., api-specs.md, rust-cpp.md) to extract doc comments and metadata.
Boundary markers: The instructions do not define specific delimiters or warnings to ignore embedded instructions within the ingested content.
Capability inventory: The skill has access to Write, Edit, and Bash tools, allowing it to modify source code and execute commands based on its analysis.
Sanitization: No explicit sanitization or filtering of external content is required before processing, although there is a constraint against including sensitive data in examples.
[EXTERNAL_DOWNLOADS]: The skill utilizes well-known documentation and linting tools through package runners.
Executes npx to download and run utilities including @redocly/cli, @stoplight/spectral-cli, widdershins, @openapitools/openapi-generator-cli, typedoc, conventional-changelog-cli, and dependency-cruiser.
Uses go run to fetch and execute the pkgsite utility from the official Go package repository.
[COMMAND_EXECUTION]: The skill performs project analysis and documentation tasks using standard command-line tools.
Uses find and ls for project structure discovery.
Uses git for commit history analysis and changelog generation.
Invokes language-specific build tools like gradlew (Kotlin), dart (Flutter), cargo (Rust), and xcodebuild (Swift) to generate API references and validate examples.

documentation-generator