Best Claude Skills for Productivity (2026)
Most "best skills" lists are copied README summaries. This one isn't: we installed all 512 productivity skills in our catalog, ran each against a baseline and scored four criteria. 143 passed cleanly, average score 8.5/10. The one to start with: Fetch URL As Markdown.
Ranked by SkillProof Score · how we score
- 1
Fetch URL As Markdown
Local trafilatura-first URL-to-Markdown fetcher with a clean exit-code contract for when to fall back to Exa.
From our test: Ran the bundled script live on a real Wikipedia page: it returned clean 44KB markdown with a proper YAML metadata header and zero nav/menu chrome, versus a 30KB naive-strip baseline still full of…
- 2
Refresh Context Map
Rebuilds the AIDEV-anchor/AGENTS.md/ADR manifest that powers claude-leverage's PreToolUse context hook.
From our test: Actually cloned the repo and ran the real script live: output matched the SKILL.md's documented format exactly ('Wrote ... (26 files, 45 anchors)') — but a fresh HEAD checkout's own committed…
- 3
Video Vision
Local ffmpeg+Whisper pipeline that lets Claude actually watch and transcribe video files
From our test: Ran the real pipeline end to end against a synthetic 6-second test clip: ffmpeg frame extraction and local Whisper transcription both executed for real and returned the exact JSON manifest shape the…
- 4
YT Digest
Turns a YouTube URL into a timestamped, screenshotted markdown research note via real yt-dlp + ffmpeg.
From our test: Ran the entire real pipeline (yt-dlp fetch, VTT parse, ffmpeg frame grab, markdown assembly, cleanup) end to end on a live YouTube video, no shortcuts: correct metadata, deduplicated captions, real…
- 5
Meeting Notes to Actions
Turns meeting transcripts into quote-cited action items - no guessed owners or due dates.
From our test: Fed it a transcript with a corrected misassignment, a joke commitment ('I'll take care of world peace'), and a vague 'someone should update the roadmap, soon': it correctly left the roadmap item…
- 6
Weekly Review
GTD-style weekly review turning wins, blockers, and OKR notes into a structured writeup.
From our test: Ran the bundled synthesizer script on a sample week: scattered wins, blockers, and OKR notes came back as a clean structured review with owner+mitigation fields per blocker and four self-coach…
- 7
NotebookLM
Scriptable Google NotebookLM CLI: notebooks, sources, podcasts, quizzes, reports, slides.
From our test: Recalling CLI commands from memory got the syntax wrong (new-notebook, add-source); following the docs produced the real commands, and they are unusually upfront about which generation types are…
- 8
Loopy
Designs, finds, audits and runs bounded AI-agent loops with evidence receipts, backed by a live Loop Library catalog.
From our test: Ran the real 'Find a published loop' workflow against the live catalog.json (85 real entries, fetched from signals.forwardfuture.com) for a flaky-CI-test scenario and it surfaced loop #044 'The test…
- 9
Critical Perspective
Reframes one-sided conversations into open questions that surface hidden assumptions and gaps.
From our test: Given the same 'go 100% paid ads, organic takes too long' pitch, the plain baseline gave a pros/cons list and a recommendation; running the skill's assumption-surfacing technique instead produced a…
- 10
Exam Drill
Generates exam-tier twin/blind/chain drills from a course you've already ingested and analyzed.
From our test: Its twin-generation rules force pattern-invariant variants with gated answer reveals, so a plain assistant asked the same thing would just print a twin problem plus the answer immediately; but the…
How this ranking works
Each skill gets a fresh install in Claude Code, three trigger phrasings, and the same task run with and without it. Scores: installation (5), trigger reliability (5), output vs baseline (10), documentation honesty (5) — normalized to a 10-point SkillProof Score. Full rubric on the methodology page. Queue-verdict skills sit at the bottom until testing completes.
New here? Start with how to install Claude skills — or grab the whole setup in one command with a curated pack ($10).
On a different AI tool? These tested instructions adapt: see our conversion guides for Cursor rules, Copilot instructions, Windsurf and Custom GPTs.