philosopher
claudePhilosopher personas: historically-grounded dialogue with source attribution, period awareness, and AI meta-cognition
24 skills
Plugin marketplace
Cognitive development plugins for Claude Code. Live experiments in AI collaboration.
Add this marketplace
Then install any plugin below by name — each page has its exact command.
Philosopher personas: historically-grounded dialogue with source attribution, period awareness, and AI meta-cognition
24 skills
Core infrastructure: git commits, model selection, hooks, scratch pad, kaizen, devil-advocate agent, context save/load, verbatim checkpoint, resumable work plans (plan-context), meta-prompt (RELIANCE recursive ledger), and session post-mortem
14 skills
Experimental and utility skills: deployment, background tasks, pinning, summarization, skill testing, distillation, and context bootstrapping
9 skills · 1 agent
OpenSpec workflow management: init, plan, design, develop, review, test, sync, reflect, replan
9 skills
Cynefin-routed problem-solving: frame, troubleshoot, investigate, brainstorm, probe, experiment, challenge, benchmark
8 skills
Business analysis toolkit: competitive analysis, UX strategy, UX wireframes, UX evaluation, UX polish, market sizing, canvas, PRD, personas
7 skills
Content transformation and writing: anonymization, infographics, literary rewriting, bridge captures, voice extraction & enforcement, deck storyline building, expense tracking
7 skills
Document and audio conversion: PDF, EPUB, DOCX, PPTX, and audio to markdown, plus markdown to PDF
7 skills
Multi-project context management: switch projects, save/load sessions, ref/wip file zones + sync, audit/reorganize project folders. Designed for non-CLI interfaces (Cowork desktop, Telegram).
5 skills
GTD workflow automation for Obsidian vaults: capture, triage, route, focus, clip
5 skills
Plugin and skill authoring: create/edit skills, agents, plugins, CLAUDE.md files, search skills, manage dependencies
5 skills
Control-plane plugin for tracking rings: recursive units of work, each a problem to settle rather than a session. A ring has a Brief frozen at open, a Trace of local working memory discarded at close, and a Delta typed by effect (ESTABLISHED, INVALIDATED, SURFACED, DELIVERABLE) that crosses exactly one turn to its parent. Ring state is orthogonal to what carries it: the four vehicles are inline, subagent, session and resumed. For vehicle session, ring layers on top of the team plugin as its vehicle — team keeps circulating mandates, reports and relay packets between named agents unchanged, while ring tracks which problem is open, who is carrying it, and what crosses back up. Ships four skills (open, close, rings, sync) for descending into and ascending out of a ring, listing/graphing the live tree, and persisting a ring's state so it survives a /clear, an on-disk writer token (carrier:) taken over by resuming a parked or active ring, exclusive create:/writes: write-right tracking with prefix-inclusion collision detection and a lineage exception for parent-child delegation, and a closing sweep that scans disk for files born or touched outside declared claims.
4 skills
Retrospective analysis: session reports, collaborative retrospectives, and domain retrospectives
3 skills
Coordination protocol for running a fleet of named Claude Code agents in parallel under one human: delegate one mandate at a time (brief), report back on a stop condition (report), hand off state at context limits (relay) — one skill per machine arrow, SendMessage the only transport. No skill reads the protocol at runtime: the mandate carries its own contract, so all three run in windows with no context to spare without reading anything at all, and a window that has never seen the plugin still behaves correctly. Context is under the human's control on both sides: read_first closes what a delegate opens before it starts, deliverable puts its findings on disk so reports and relay packets carry pointers rather than content, and a relay hands its successor the role lines that land it in READY instead of straight back into the work.
3 skills
Visual output an agent's reader grasps at a glance. Ships the Digital Stoic design system v0.6 "Tablette" (design-system): tokens, type, class vocabulary, inline-SVG chart parameters, and an assembler that injects the base CSS and base64 fonts so they never pass through context. Six demo screens (document, landing, dashboard, charts, decision, mini-app) kept in sync by a checker. Also ships verdict (ex review-html): self-contained review pages assembled from the design system, with cards typed by the action the reader takes and a JSON of deviations pasted back, plus its 108-entry visual grammar catalog and generator.
2 skills
Personal coaching with CLEAR protocol, anxiety-first surfacing, and Goldsmith accountability loop
1 skill
Lazy skill demand capture: declare placeholder skills that intercept trigger keywords and log hits to measure real demand before building the full implementation. Toggle off to pause capture.
1 skill
Sandbox for Claude Code mods experiments (not published)
Next plugin generation (experimental), replaces dstoic. Ships the pack mod: /pack saves a stream's progress and arms a fresh start after /clear, /pack save only saves, /pack cancel drops an armed fresh start, /unpack <stream> loads a saved stream in a new session.
Shows how many plugins this repo ships, and links here.