Recently added

Sep 23, 2026

1. cobik0stars

   Skill ของ cobik ที่ติดมากับแอป Cobik Desktop — ทุกคนได้เหมือนกันโดยไม่ต้องติดตั้งเอง

   Claude Code6 skills
2. orca170stars

   Delegate a well-defined implementation task to Orca's autonomous plan-code-review flow.

   Claude Code1 skill
3. tokamak3stars

   Build, run, debug, and design applications with tokamak.

   CodexClaude CodeAgent Plugins1 skill
4. tekton0stars

   tekton: author your own project content as valid Autodesk Revit (.rvt/.rfa) files with no Revit install and no base file — from a prompt, an IFC, or an existing .rvt. Create rooms, walls, panelboards, families and circuits; edit, move, retype and delete elements; validate and inspect any file. A licensed engineer opens the output in Autodesk software for final QA.

   Claude Code4 skills
5. tekton0stars

   tekton: author your own project content as valid Autodesk Revit (.rvt/.rfa) files with no Revit install and no base file — from a prompt, an IFC, or an existing .rvt. Create rooms, walls, panelboards, families and circuits; edit, move, retype and delete elements; validate and inspect any file. A licensed engineer opens the output in Autodesk software for final QA.

   Claude Code4 skills
6. aid-orchestrator3stars

   AID — AI Development Orchestrator. 6-state bash FSM, Fast Mode (<2 min overhead), 87% token reduction. Takes a task or EPIC, generates a Plan, dispatches role-based agents, enforces quality gates, and produces auditable evidence.

   Claude Code
7. x0stars

   Dima's own skills — code core values + language guides, CST session handoff, commit workflow, RL task authoring, prompt review, agentic loops, pm ticket quick-ops, cross-repo conventions

   Claude Code25 skills
8. x-cw0stars

   the x skill set curated for cw (claude desktop) — symlinks into plugin-x, plus cw-only skills as real dirs

   Claude Code2 skills
9. cclio0stars

   cclio's coordinator rituals — boot, status board, flawlog habit, halt. typed deliberately, never self-triggered.

   Claude Code6 skills
10. devio0stars

    How a Devio designer works with Claude Code: research before any task, follow the community's standard, and run engineering through mattpocock-skills.

    Claude Code
11. agnostic-ai16stars

    One spec, every AI CLI. Install the CLI, scaffold specs, import an existing config, and keep every tool's generated files in sync.

    Claude Code4 skills
12. alex-coding0stars

    Project-context-driven setup, planning, implementation and PR review/feedback monitoring

    CodexClaude Code5 skills
13. hopkin0stars

    Analyze ad performance, email marketing, search console, and web analytics data across Google, Meta, LinkedIn, Reddit, TikTok, ChatGPT Ads, Mailchimp, Google Search Console, and Google Analytics directly in Claude using Hopkin MCPs

    Claude Code9 skills9 MCP servers
14. policy-graph441stars

    Navigate cnspec policy bundles using graph commands for LLMs

    Claude Code1 skill
15. mql441stars

    MQL query development with syntax guidance, platform-specific patterns, and MCP tool integration

    Claude Code1 skill
16. cnspec-skills441stars

    cnspec agent skills for MQL development and policy navigation

    Claude Code2 skills
17. rhiza4stars

    Slash commands for rhiza-managed repos: init and sync from a pinned template release, score code quality, refresh README/CLAUDE.md/mkdocs, prepare releases, and report sync status.

    Claude Code11 skills
18. kpopper4stars

    Make your agents' reasoning explicit, persistent, and deterministically checkable. Connect decisions to evidence and assumptions, and see what needs another look when a recorded premise changes.

    CodexClaude Code10 skills
19. pair-watch0stars

    Supervised pair for coding agents, launched from one chat: the invoking chat is the read-only watcher and starts implementer sessions itself (tmux), verifies their reports against git, tests and session logs, and runs the review gates. Claude seats via SendMessage; Codex seats via native tmux launch, queue notifications and sequenced files, including Codex + Codex.

    CodexClaude Code1 skill
20. harness0stars

    Enforcement layer for AI-assisted development: hooks, gates, and evals that make the workflow mandatory, not advisory

    Claude Code3 skills
21. md-eval0stars

    eval-docs skill: score Markdown docs for AI coding agent readiness and human readability with the md-eval CLI (TypeSafe Jev).

    Claude Code1 skill
22. md-eval-plan-gate0stars

    Scores each plan with md-eval before Claude leaves plan mode and sends weak plans back once with the gaps to fix.

    Claude Code
23. narness0stars

    Harness engineering: constrain AI agents with code, hooks and scripts instead of prompts — skills, validation scripts, git hooks and per-tool configs

    Claude Code1 skill
24. workflowy0stars

    Workflowy 노드 하나를 작업 흐름(workstream)의 기록으로 삼아, Claude Code 세션들이 그 아래에 작업 과정을 이어서 정리해 기록한다.

    Claude Code1 skill1 MCP server
25. pseudoplan0stars

    Two modes, one review page. Plan a change as pseudocode and get it approved before any code is written, or map an existing diff into the same diagram, with the real code attached, to review someone else's change. Feedback goes straight back to the agent.

    Claude Code2 skills
26. shapeup-sdlc-plugin3stars

    Shape Up SDLC harness for Claude Code: shaping, intake, orient, scope-mapping, building (T0-verified, sandboxed, scope-contracted), evaluation and QA skills orchestrated by a tech-lead.

    Claude Code12 skills
27. searchlink-lite0stars

    Google Search Console reports in Claude, Cursor and other MCP clients. Local and read-only.

    Claude Code1 MCP server
28. vaadin-agent-tools2stars

    Vaadin tools for AI agents: inspect and validate Vaadin projects with a self-contained native CLI.

    CodexClaude Code2 skills
29. bgutschke-skills0stars

    Personal agent skills for real engineering.

    Claude Code
30. watermarks-remover0stars

    Remove multi-vendor AI provenance marks: invisible Unicode (Layer A), statistical text watermarks via rewrite (Layer B), and C2PA/EXIF/XMP/container metadata on images, documents, and media. Ships the remove-ai-marks skill (thin client over the repo's HTTP service) and the self-contained clean-user-facing-text skill.

    Claude Code
31. bend-spec0stars

    Spec-first work for Bend projects traced by bolt (ez, bolt, eztoml, ezhttp, ezjson, snap, shake): bringing a project under spec, and adding features to one that is.

    Claude Code2 skills
32. skill-architect0stars

    Agent skills for auditing and improving Agent Skills against the spec, Anthropic best practices, and ICM context-management criteria.

    CodexClaude Code2 skills
33. neuraldsp-preset-generator0stars

    Generate, match, and edit Neural DSP amp-sim presets from a song, reference recording, or plain-English change. Personal interoperability with your own licensed copy of the plugin.

    Claude Code3 skills
34. product-handover0stars

    プロダクトをエージェントに引き渡すときにだけ必要になる固有部品を持つプラグイン。法務ドラフト（利用規約・プライバシーポリシー・返金ポリシー）の雛形、サポート窓口メールの設定雛形、教訓ログの雛形の3つだけを配る。インフラ構築・開発ワークフロー導入・auto-merge 配線・SNS 運用はそれぞれ infra / dev-workflow / sns-autopilot が持つため本プラグインは呼び出さず、導入順だけを README に書く。住人側（workspace・cron・チャンネル）は flatmate の new-resident が担い、本プラグインはプロダクトのリポ側だけを扱う。

    Claude Code
35. cost-ledger0stars

    issue / PR 単位の API 換算コストを、Claude Code の会話ログ（`${CLAUDE_CONFIG_DIR:-$HOME/.claude}/projects/**/*.jsonl`）から集計する。帰属はブランチと（リポジトリ識別子, issue 番号）の 2 本立てで、サブエージェント（isSidechain）の消費も同じブランチに寄せる。`/cost <番号>` で PR ならヘッドブランチの総額、issue なら投稿を境界に切った区間の合計を USD と円で返す。pr-review-gate が `agent-review:passed` を付けた直後に、PostToolUse の hook がその PR へ `/cost` の 1 行目を 1 本のコメントで貼る（環境変数 `COST_LEDGER_GATE_REPORT=off` で止まる）。実行時に python3（3.8 以上）と、番号の判別に認証済みの gh を必要とする。git は 2.31 以上（rev-parse --path-format=absolute を使う）。

    Claude Code
36. fusebase6stars

    Comprehensive FuseBase MCP server and agent platform tool suite for workspaces, Y.js live collaboration, dashboards, visual tables, CRM pipelines, client portals, automations, and Gate PostgreSQL stores.

    CodexClaude Code1 MCP server
37. huitzo0stars

    Build Intelligence Packs and Dashboards on the Huitzo platform: SDK, manifest, CLI, platform API/MCP and dashboard SDK references, docs-first workflow skills, developer/reviewer agents, safety hooks and a project-docs MCP server.

    Claude Code
38. protocol-guard0stars

    Required workflow checks with Claude Code function hooks. Refuses a knowledge-file write until knowledge-save is open, refuses hand edits to generated indexes, refuses a pull request, close or merge until knowledge-save (and the work or merge-and-clean-up skill) is open, and holds a final reply once when working memory or the index rebuild and checker did not follow the change. Fact checks only: no model call and no reading of the agent's words. Runs only when CLAUDE_CODE_ENABLE_FUNCTION_HOOKS=1.

    Claude Code
39. wonderpet-general3stars

    萬達寵物通用 AI 工具包. Includes 7 skills.

    Claude Code7 skills
40. ragdoll-workspace3stars

    Ragdoll 工作區插件. Includes 4 skills.

    Claude Code4 skills
41. norwegianforest-workspace3stars

    NorwegianForest 工作區插件. Includes 10 skills.

    Claude Code10 skills
42. maltese-workspace3stars

    Maltese POS 工作區插件. Includes 6 skills.

    Claude Code6 skills
43. questionpunk0stars

    Create and review survey drafts, publish requested studies, and analyze QuestionPunk results.

    Claude Code4 skills1 MCP server
44. dev-lead2stars

    Cross-model delegation and adversarial review for CLI coding agents: one lead, six runtime adapters, and no change merges reviewed only by its own model family.

    Claude Code14 skills
45. qa-kit0stars

    QA toolkit: clarify and solve problems (grill-me, brainstorm, problem-solving, predict), generate test cases and edge cases, write Playwright/Vitest/k6 automation, run and audit test suites, explore apps in a browser, and debug failures to root cause.

    CodexClaude Code9 skills
46. gecko-surf7stars

    Make any API agent-usable — even one with broken docs, no OpenAPI, a WAF, or an auth wall. Comprehend a spec (or JS-rendered docs) into first-call-correct MCP tools your agent calls right the first time, with auth injected and no integration code. On a painful real API, first-call-correctness goes from 10% to 65% (measured); wire x402 pay-per-call; defend against poisoned specs. Installs a live demo surface (TxODDS) over MCP — try it with zero setup.

    Claude Code1 MCP server
47. gecko-surf7stars

    Comprehend any API or Solana program into first-call-correct agent tools, and check the call before it counts: simulate the exact bytes, predict compute, refuse on stale evidence.

    Agent Plugins5 skills
48. gecko-let-me-buy7stars

    Buy from a Let Me Buy storefront on Solana and confirm delivery, with the traps that the program's IDL does not state.

    Agent Plugins2 skills
49. trajectoryeval0stars

    Trajectory-match verifier sibling: checks the PATH a worker took (its ordered tool-call sequence) against an expected trajectory spec, in strict/unordered/subsequence modes. v0.1.7: adds a tolerance-based `fuzzy_hash` diff strategy — `TierConfig::threshold_permille` config plus a pure `fuzzy_diff()` computing a deterministic permille drift distance (differing leaves / total leaves) that tolerates drift at or under the threshold and escalates to `DiffOutcome::DriftedBeyondThreshold` above it; distance is stored as `u32` to keep the `Eq` derives on `DiffOutcome`/`TierVerdict` intact (no `f64`). v0.1.6: the seeded 1-in-N sampling hash now delegates to `harness_core::hash::fnv1a64` (the single canonical FNV-1a implementation) instead of a private reimplementation, closing a code-duplication finding; hash values are unchanged (same algorithm/constants), so sampling behavior is bit-for-bit identical. v0.1.5: `tier` subcommand is now e2e-tested end-to-end via a committed example core-flow allowlist (examples/tier-config.json); seeded 1-in-N non-core sampling has a rate-pinning test (catches subtler mutations than mere non-constancy); a core flow using the unimplemented `screenshot` diff strategy now yields an explicit tri-state `needs-human` verdict (distinct exit code 3) instead of silently masquerading as a hard diff failure. v0.1.4: adds risk-tiered e2e verification (`tier` subcommand) — a config-driven core allowlist where core flows get an every-run deterministic structured-data snapshot diff (perceptual-hash/screenshot behind a documented stub boundary) and non-core flows get an existence check or seeded low-frequency sampling. Subscription-native (one bundled Rust binary, no API key).

    Claude Code
50. tracekit0stars

    Span-tree tracer for condukt runs: record one run's interpreter→worker→verifier phases as a parent-linked span tree (phase, model, ms, cost, status), render it with `tracekit trace <RID>`, and export OpenTelemetry GenAI-semconv JSON. Answers which phase of a failed run was slow/expensive/broke. Subscription-native (one bundled Rust binary, file-only, no network, no API key).

    Claude Code