Agent Plugins Marketplace

Recently added

Plugins grouped by the day they were added to the directory, newest first.

Aug 25, 2026

  1. foursday15stars

    Operate the local Foursday work twin from Claude Code through the shared Control MCP.

    Claude Code1 skill1 MCP server
  2. trazum1stars

    Price a prompt before it costs money, and read the bill after. Deterministic prompt optimisation (no API key, no network), token budgets and drift gates for CI, usage-log profiling that names what would actually move the bill, and seven MCP tools including a spend guard that judges configured USD ceilings before a call is made.

    Claude Code1 skill1 MCP server
  3. Gmail MCP server for inbox triage: search with re-verified results, labels, archive/trash, mailbox-side snooze, triage digest, RFC 8058 unsubscribe. No send tools by design. Runs locally via npx; one-time `npx mailwarden --auth` connects your account.

    Claude Code1 skill1 MCP server
  4. agent-kit0stars

    Bộ điều phối subagent cho Claude Code với phê duyệt kế hoạch theo prompt, biên nhận kiểm chứng có cấu trúc và kiểm tra glossary khớp chính xác.

    Claude Code1 skill
  5. Checks whether an AI coding agent's claims (tests passed, file created, build succeeded, commit made) are actually true, using independently captured evidence instead of trusting its self-report.

    Claude Code
  6. Has marketplace

    Codex
  7. skills-no-fm757stars

    Skills without frontmatter

    Codex
  8. Declares skills but dir missing

    Codex
  9. my-plugin757stars

    A plugin with no version

    Codex
  10. beta-plugin757stars

    Beta plugin with intentionally incomplete metadata.

    Codex1 skill
  11. Codex package in mixed repo

    Codex
  12. my-plugin757stars

    my-plugin is an open-source Codex plugin indexed from GitHub. Review its manifest and source before installing.

    Codex
  13. A minimal plugin

    Codex
  14. Fixture with a dangerous local skill definition

    Codex
  15. A well-structured example plugin. Includes 1 skill.

    Codex1 skill
  16. Example Claude plugin package

    Claude Code
  17. pstack0stars

    if you want to go fast, go deep first. pstack helps you write less, but higher quality code. rigorous agent workflows you can parallelize with confidence. Claude Code port of cursor/plugins/pstack.

    Claude Code44 skills
  18. slash0stars

    Skills et automatismes de traitement des tickets Linear SLI : parcours du ticket, constat partagé, jeu de données de recette, taille et découpage des pull requests, rédaction des écrits lus par un humain — écrits GitHub et livrables longs, ancrage du navigateur.

    Claude Code2 skills1 MCP server
  19. mattpocock-skills is an open-source Claude Code plugin indexed from GitHub. Review its manifest and source before installing.

    Claude Code
  20. mattpocock-skills is an open-source Claude Code plugin indexed from GitHub. Review its manifest and source before installing.

    Claude Code
  21. mh3stars

    A Claude Code harness that composes mattpocock-skills instead of duplicating it. Nine computational deny/ask gates (irrecoverable commands, subagent git guard, subagent spawn guard, task-complete separation, test integrity, config-write guard, secret-content scan, codex-setup guard, routine-trigger guard) plus a SubagentStop verdict-check gate, a 4 KB METHODOLOGY injected at SessionStart, and memory-lint for the auto-memory store.

    Claude Code
  22. quorum0stars

    A four-step delivery pipeline (plan, build, multi-lens review, adjudication) with behavioral regression tests and CI gating.

    Claude Code7 skills
  23. Personal, agent-agnostic collection of reusable skills.

    Claude Code6 skills
  24. brain-dev0stars

    Rootcause Brain Dev skills for local checks, safe cross-computer Git sync, rc ask validation, production-run debugging, health/fleet review, and exact-SHA publish workflows. Brain-dir-relative; no private RootCause source needed.

    Codex12 skills
  25. signal1stars

    Plain English harness. States the budget before the agent writes, and reads what came back.

    Claude Code3 skills
  26. pest1stars

    Pest testing suite skills. Philosophy, organization, factories, datasets, arch testing, mocking decisions.

    Claude Code11 skills
  27. panel1stars

    Adversarial agent teams. Verification that cannot be self-issued. Standalone; pairs well with kernel.

    Claude Code5 skills
  28. The package author's playbook. Ecosystem-wide patterns for crafting, shipping, and sustaining pinnacle Laravel packages.

    Claude Code29 skills
  29. Laravel DDD patterns. Opinionated domain-driven design with Actions, DTOs, QueryBuilders, and decision heuristics. Standalone; pairs well with kernel.

    Claude Code47 skills
  30. kernel1stars

    Universal cognitive kernel. Stack-agnostic thinking patterns, memory management, and planning tools. The foundation for all Craftsman plugins.

    Claude Code39 skills
  31. floor1stars

    Where work happens. A run is one attempt at one work item, held outside every repository it changes.

    Claude Code1 skill
  32. wenlan51stars

    A living knowledge base your agents build as they work. What they learn becomes source-cited wiki pages that refresh between sessions, so each new thread starts from your latest work.

    Claude Code11 skills1 MCP server
  33. wenlan51stars

    A living knowledge base your agents build as they work. What they learn becomes source-cited wiki pages that refresh between sessions, so each new thread starts from your latest work.

    Codex11 skills1 MCP server
  34. core1stars

    core is an open-source Claude Code plugin indexed from GitHub. Review its manifest and source before installing.

    Claude Code
  35. execution-observation is an open-source Claude Code plugin indexed from GitHub. Review its manifest and source before installing.

    Claude Code
  36. The orchestration loop for contract v3: draft issues from the user's words, spawn role sessions (core plugins + role rulebook, agent account, issue-<n>/<role> branch), explain the PRs that come back, and relay the user's decisions — comments, review Approve, merge — with the user's own account, per conversation.

    Claude Code
  37. The ux-engineering role on contract v3: 디자인 결정 → 토큰/규칙 시스템화. use_when: 화면 스펙이 여러 개 쌓여 시스템화가 필요할 때. Hand-off: 브랜드 정체성 결정이 필요하면 → brand-design; 접근성 기준 미달이면 → accessibility.

    Claude Code
  38. The technical-writing role on contract v3: 독자가 알아야 할 것을 어떻게 구조화할지. use_when: 외부 공개 문서가 필요할 때. Hand-off: 개발자 대상 온보딩이면 → devrel.

    Claude Code
  39. The security-threat-model role on contract v3: 신뢰 경계의 위협 표면. use_when: 스펙에 신뢰 경계·인증·민감데이터가 걸릴 때. Hand-off: 구현 단계 취약점 점검은 → secure-coding; 법적 노출이면 → legal-compliance.

    Claude Code
  40. The secure-coding role on contract v3: 구현이 공격에 견디는가. use_when: 인증/입력처리 코드 랜딩 후. Hand-off: 설계 단계 위협표면 재검토가 필요하면 → security-threat-model.

    Claude Code
  41. sales1stars

    The sales role on contract v3: 리드/기회를 어떻게 진행시킬지. use_when: 영업 프로세스 설계가 걸릴 때. Hand-off: 메시지/포지셔닝 자체는 → marketing.

    Claude Code
  42. The risk-management role on contract v3: 전사 리스크 노출이 허용 범위인가. use_when: 재무/운영/전략 리스크가 걸릴 때 (feasibility보다 넓은 범위). Hand-off: 개별 법규 컴플라이언스 세부는 → legal-compliance.

    Claude Code
  43. pricing1stars

    The pricing role on contract v3: 얼마를, 어떤 구조로 받을지. use_when: 신규 가격 정책이 걸릴 때. Hand-off: 단위경제 성립 여부 재확인은 → finance-unit-economics.

    Claude Code
  44. The pr-communications role on contract v3: 메시지가 외부에 어떻게 읽힐지. use_when: 외부 커뮤니케이션이 걸릴 때. Hand-off: 캠페인 성격 메시지는 → marketing.

    Claude Code
  45. The performance-engineering role on contract v3: 부하/지연 목표를 만족하는가. use_when: 성능 예산이 걸린 설계/회귀일 때. Hand-off: 용량 증설 타이밍이 걸리면 → capacity-planning.

    Claude Code
  46. The partnerships-bd role on contract v3: 파트너십이 구조적으로 성립하는가. use_when: 제휴/BD 딜 구조가 걸릴 때. Hand-off: 법적 계약 검토는 → legal-compliance.

    Claude Code
  47. The ml-engineering role on contract v3: 모델을 서비스로 안정적으로 서빙 가능한가. use_when: 모델 서빙 표면이 걸릴 때. Hand-off: 학습 데이터 파이프라인이면 → data-engineering.

    Claude Code
  48. marketing1stars

    The marketing role on contract v3: 어떤 메시지로 어떤 채널에 도달할지. use_when: 캠페인/포지셔닝이 걸릴 때. Hand-off: 퍼널 성과 해석은 → growth-analytics.

    Claude Code
  49. The market-analysis role on contract v3: 경쟁 구도에서 이 스펙이 서는가. use_when: product 스펙 확정 후, 경쟁 구도가 걸린 결정일 때. Hand-off: 가격 정책이 걸리면 → pricing; 포지셔닝 메시지가 걸리면 → marketing.

    Claude Code
  50. The localization role on contract v3: 다른 로케일에서도 산출물이 성립하는가. use_when: i18n 대상 표면이 걸릴 때. Hand-off: 카피 원문 자체를 다시 써야 하면 → content-design.

    Claude Code