using-rtfd

name	using-rtfd
description	Use PROACTIVELY when writing or modifying code that calls external library APIs — verify function signatures, parameters, and return types are current before writing the code. Also use when looking up library docs, package metadata, Docker images, GitHub repos, GCP services, LogScale query syntax, or Zig docs.

Using RTFD

RTFD provides 33 tools across 10 providers for real-time documentation lookup. Core pattern: search first, then fetch full docs.

When to Use

Proactive (use without being asked)

• Writing code that uses external libraries — look up current API before writing calls, especially for libraries that change frequently or that you're not 100% certain about
• Generating install commands or dependency specs — verify the current version exists
• Recommending a library or tool — confirm it's still maintained, check latest version
• Unsure about a function signature, parameter name, or default — fetch the docs rather than guessing

Reactive (user asks for help)

• Need current docs for any library/package (PyPI, npm, crates.io, Go, Zig)
• Looking up package metadata, versions, or stats
• Browsing GitHub repos (tree, files, diffs, code search)
• Finding Docker images or inspecting Dockerfiles
• Querying GCP service documentation
• Writing LogScale/Humio queries (syntax, functions, operators)

Quick Lookup

Search (start here)

Tool	Searches
search_library_docs	All ecosystems at once (PyPI + npm + crates + GoDocs + GitHub)
github_repo_search	GitHub repositories (filter by language)
github_code_search	Code patterns (optionally scoped to repo)
search_crates	crates.io
search_docker_images	DockerHub
search_gcp_services	Google Cloud Platform services
search_logscale_docs	LogScale syntax, functions, operators

Metadata (quick info)

Tool	Returns
pypi_metadata	Version, URLs, project links
npm_metadata	Version, repository, maintainers
crates_metadata	Version, docs link, download stats
godocs_metadata	Package summary, source URL
docker_image_metadata	Stars, pulls, official status, tags

Fetch Full Docs

Tool	Source
fetch_pypi_docs	PyPI README (RST auto-converted to Markdown)
fetch_npm_docs	npm README
fetch_godocs_docs	godocs.io / pkg.go.dev
fetch_gcp_service_docs	cloud.google.com
fetch_github_readme	GitHub repo README
fetch_docker_image_docs	DockerHub README
fetch_dockerfile	Dockerfile from source repo (best-effort)
zig_docs	Official Zig language docs

GitHub Repo Exploration

Tool	Purpose
get_repo_tree	Full file tree (set recursive=True)
list_repo_contents	Directory listing at a path
get_file_content	Read a specific file (up to 100KB)
get_commit_diff	Diff between commits, branches, or tags
list_github_packages	GHCR packages for an owner
get_package_versions	Package version history

LogScale Query Reference

Tool	Purpose
list_logscale_functions	Browse functions by category
logscale_syntax	Syntax topic docs (filters, regex, time, macros...)
logscale_function	Single function signature and docs

Admin

Tool	Purpose
get_cache_info	Cache stats (entry count, size)
get_cache_entries	Detailed cache entries with age/preview
get_next_chunk	Retrieve next chunk when response was truncated

Common Workflows

Library docs (any ecosystem): search_library_docs("name") → pick ecosystem → fetch_pypi_docs / fetch_npm_docs / etc.

GitHub exploration: github_repo_search("query") → get_repo_tree("owner/repo") → get_file_content("owner/repo", "path")

Docker investigation: search_docker_images("query") → docker_image_metadata("image") → fetch_docker_image_docs("image") → fetch_dockerfile("image")

LogScale query help: search_logscale_docs("topic") → logscale_syntax("topic") or logscale_function("name")

Key Details

• Repo format: Always "owner/repo" (e.g., "psf/requests")
• Chunking: Large responses include a continuation_token — call get_next_chunk to get more
• max_bytes: All fetch tools accept this param (default 20480) to control response size
• Caching: Results cached for 1 week by default
• search_library_docs is the universal entry point — use ecosystem-specific tools only when you already know the ecosystem