tldr-pages/tldr

Collaborative cheatsheets for console commands 📚.

Markdown62699 starsEducation & ReferenceGitHub

Quality: solid 18/24

PAI: integrate 0.63

Overview

What it is — tldr-pages is a community-maintained collection of simplified, example-first help pages for thousands of command-line tools across Linux, macOS, Windows, BSD, and Android.
Problem — Traditional man pages are verbose and often lead with obscure edge cases rather than the common usage patterns most users actually need.
Who it's for — Any command-line user — beginner or expert — who wants a quick, practical reference instead of wading through dense manual pages.
Notable — With 62K+ stars, daily commits since 2013, and coverage spanning 10,000+ pages across multiple platforms and client implementations, it is the dominant open standard for community CLI documentation.

Verdict

	Rating	Summary
Quality	solid (18/24)	Extremely active, well-documented reference project with exceptional adoption; license metadata ambiguity and stale release cycle are minor marks against it.
PAI Relevance	integrate (0.63)	The tldr CLI client is a clean subprocess target filling a gap PAI has no coverage for — quick command-syntax lookup — and the example-first page format is worth studying for PAI's Knowledge subsystem organization.

Quality Assessment

18/24 — maintained / adequately-documented / high-discipline

Health: 5/8 (maintained)

Failed:

H2: FAIL — Latest release v2.3 was 2025-03-07, approximately 14.7 months before appraisal date (>12-month threshold)
H6: FAIL — 228 open issues exceeds the healthy ceiling of <100
H7: FAIL — License field reports NOASSERTION in metadata; GitHub's detection failed despite CC BY 4.0 badge in README

Passed:

H1: PASS — Release v2.3 exists (2025-03-07)
H3: PASS — Last commit 2026-05-28 (same day as appraisal)
H4: PASS — Active push activity within 30 days
H5: PASS — isArchived: false
H8: PASS — README contains GitHub Actions CI badge linking to .github/workflows/ci.yml

Documentation: 6/8 (adequately-documented)

Failed:

D5: FAIL — No heading matching API, Configuration, Options, Reference, Commands, or Parameters; client sections use prose headings like "Python client" rather than a structured command reference
D8: FAIL — No Limitations, Caveats, Known Issues, or Trade-offs section present in the README

Passed:

D1: PASS — README is present and non-empty
D2: PASS — README is substantially longer than 1000 bytes (8KB provided alone)
D3: PASS — Explicit install instructions for Python (pipx install tldr), Rust (brew install tlrc, cargo install tlrc), and Node.js clients
D4: PASS — Multiple code blocks under usage headings showing tldr tar, man tar contrasts, and console examples
D6: PASS — Purpose stated in first paragraph: "a collection of community-maintained help pages for command-line tools, that aims to be a simpler, more approachable complement to traditional man pages"
D7: PASS — Links to https://tldr.sh/, wiki, PDF release, and Matrix chatroom

Engineering Signals: 7/8 (high-discipline)

Failed:

E1: FAIL — Primary repository language is Markdown (content repo); clients are separate repos in Python/Rust/Node.js

Passed:

E2: PASS — package.json present with lint/test/build scripts
E3: PASS — Only 3 direct dependencies (glob, markdownlint-cli, tldr-lint), well within threshold
E4: PASS — manifest includes "test": "bash scripts/test.sh" and CI references test pipeline
E5: PASS — 62,699 stars, far exceeding 50-star threshold
E6: PASS — ~418 stars/month (62,699 stars over ~150 months since Dec 2013)
E7: PASS — 5,255 forks
E8: PASS — Description "Collaborative cheatsheets for console commands 📚." is clear and meaningful

PAI Relevance

Dimension	Score	Assessment
Harvest Value	1	The tldr page format — platform-tagged, example-first, one-sentence-summary structure — is an interesting knowledge-organization pattern worth studying for PAI's Knowledge subsystem, which currently stores free-form markdown rather than structured command schemas.
Integration Readiness	1	No TypeScript client exists; the Python (`tldr`) and Rust (`tlrc`) CLI clients are both installable and produce stdout text suitable for subprocess wrapping in a PAI skill, but adapter code is required.
Overlap Risk	0	PAI has Research, Knowledge, and ContextSearch skills but none perform quick CLI command-syntax lookup; this fills a distinct gap not covered by any existing capability.
Gap Fill	1	PAI agents frequently construct shell commands and could benefit from a fast, offline-capable reference tool; no existing skill or tool in the manifest addresses command-syntax lookup specifically.

Composite: 0.63

What Next

fab (Fabric Recommender) subprocess construction: When fab builds a command pipeline around a recommended Fabric pattern, it currently relies on the LLM's training knowledge for flags and syntax of tools like jq, fzf, or xargs. Add a tldr_lookup step that fetches https://raw.githubusercontent.com/tldr-pages/tldr/main/pages/common/<tool>.md and injects the examples block into context before any subprocess call involving a tool not in a pre-approved list. Outcome: fab constructs syntactically valid pipelines for unfamiliar tools on the first attempt rather than hallucinating option names.
Capture-to-Knowledge Pipeline validation: When the pipeline encounters shell invocations in captured content (tutorials, blog posts, READMEs), the current validation step passes raw or training-derived syntax to the LLM. Replace that with a tldr lookup as the first-pass reference — the pages are plain Markdown, fit easily under 500 tokens, and cover 10,000+ tools across Linux/macOS/Windows. Outcome: Validation prompts for CLI-heavy captures stay within context budget and ground the LLM in actual documented usage rather than hallucinated variants.
Agent system — universal subprocess pre-flight: Across Conservancy deploys (Wrangler, git, Cloudflare CLI) and pipeline orchestration scripts, agents constructing shell commands have no reliable compact reference source today. Create a single tldr_fetch(tool, platform="common") utility that resolves the raw GitHub URL and returns just the ## Examples section. Wire it as a pre-flight check before any agent-issued subprocess call where the tool was not already seen in the session. Outcome: A reusable, zero-dependency lookup that works in any project context without installing a tldr client binary.

Landscape Position

Category: Education & Reference

In this category: codecrafters-io--build-your-own-x, practical-tutorials--project-based-learning, public-apis--public-apis, Shubhamsaboo--awesome-llm-apps, HenryNdubuaku--maths-cs-ai-compendium (from landscape context; none have prior formal appraisals, so crowding_index is 0)

Standing: By star count (62K+) and activity (daily commits, 12+ year history), tldr-pages is unambiguously the dominant reference resource in this category — larger and more active than any other repo currently tracked here.

Evidence Base

Density: 9/10 — Available: repository metadata (stars, forks, dates, language, license), full README (8KB), dependency manifest (package.json), topics list, release data. Missing: actual page content samples beyond README illustrations, CONTRIBUTING.md detail, CI workflow file, full open issues breakdown.

Notes

The license field NOASSERTION is a GitHub detection artifact — the project is explicitly licensed CC BY 4.0 as shown in the README badge and LICENSE.md reference. Health probe H7 fails on the raw metadata but the repo is not actually unlicensed. The 228 open issues for a repo this large likely reflects a healthy PR/contribution queue rather than neglect — the project merged thousands of PRs and commits daily. Integration in PAI context means installing tlrc or tldr as a subprocess tool, not importing the markdown corpus directly.