Anthropic's official meta-skill for drafting new Claude skills, testing them against baselines, and tuning descriptions for reliable triggering.
Official skill that appends 2–3 targeted "worth a second look" questions after answers the user might act on.
Dump your messy stream of thoughts, tasks and ideas and get back a clean four-section system (Projects/Ideas, Tasks, Connections, How I Can Help) with zero information loss.
A scenario-analysis skill that expands any "what if?" question into 4–6 branches (best, likely, worst, wild card, contrarian, second-order) with probabilities, trigger signals and response plans.
Summons 3–6 of 12 distinct thinking archetypes to argue over your question, then synthesizes the tension into an actionable recommendation.
Reshapes Claude's replies so the next action comes first, multi-step work is numbered, and progress is restated every turn.
An output-style skill that reshapes Claude's answers to lead with the next action, number steps, and drop all filler.
A skill that compresses images to WebP, PNG, or JPEG using whichever tool is available on your machine.
Pull YouTube transcripts (multi-language, translated, SRT), chapters, speaker labels and cover images from just a URL or video ID.
Fetch any webpage through Chrome and save it as a clean Markdown file.
A meta-skill that turns a directory, URL, pasted notes, or the work you just did into a reusable SKILL.md — or refines an existing one in place.
A blunt Marc Andreessen persona that pressure-tests ideas and ventures market-first and ends with a BUILD / DERISK / KILL verdict plus confidence level.
Conservatively detects the CPU, memory, disk and accelerator limits your current process actually has, emitting a redacted JSON snapshot plus safe worker-planning helpers.
Reads any text and maps the author's thinking style across 12 cognitive dimensions into a 'cognitive fingerprint' profile.
Auto-detects any knowledge source — docs site, GitHub repo, PDF, video, codebase — and turns it into a packaged AI skill.
A conversational package manager for Claude Code skills — search, install, update and remove skills from the CCPM registry without typing CLI commands.
Turns any image path or URL into structured JSON (OCR, layout, semantics) so text-only models can 'read' images.
A prompt-engineering skill that detects your target AI tool and returns one paste-ready prompt tuned to that tool's quirks.
A safety net that preserves your task state and key files across Claude Code context compaction cycles.
A practical playbook for managing token budgets and context quality in Claude sessions through four operations: Write, Select, Compress, Isolate.