Agent Plugins Marketplace

Recently added

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

Sep 22, 2026

  1. acceptance0stars

    Verify features and fixes against requirements, collect product evidence, and publish results for review with LobeHub.

    Claude Code1 skill
  2. qa-ai-stlc0stars

    Open-source, model-agnostic QA framework: explorer, test design, execution and reporting.

    Claude Code3 skills1 MCP server
  3. pm0stars

    Product discovery to PRD for product managers: research public sources and choose an opportunity (/pm:scout), build a clickable prototype (/pm:prototype), write a repository-grounded PRD with an independent read-only review (/pm:prd), and turn one of your own PM practices into a Claude skill (/pm:create-skill).

    Claude Code4 skills
  4. bios-implant0stars

    Agent University BIOS and worldmodel plugin for Codex with portable skills and a local companion MCP.

    CodexClaude Code5 skills2 MCP servers
  5. au-cowork-personal0stars

    Room-centric collaboration over the shared ours.network daemon.

    CodexClaude Code1 MCP server
  6. au-cowork-moderator0stars

    Moderator room tools using the assigned existing identity.

    CodexClaude Code1 MCP server
  7. ship-flow0stars

    Delivery-loop skills — land and release already-verified work through a repo's real gates. Tracker- and branch-model-agnostic: reads a consuming repo's own config instead of hardcoding one repo's setup.

    Claude Code30 skills
  8. loop-memory0stars

    pgvector semantic recall for verified dev-loop lessons (and, optionally, ADR/glossary/research knowledge). Opt-in: installs disabled by default — this is a database dependency, not core loop mechanics.

    Claude Code
  9. loop-engine0stars

    Core loop mechanics: verdict-run (machine-readable PASS/FAIL contract), loop-fix (closed verify->fix loop), lessons (verified-fix memory), classify-risk, require-tests, check-docs-hygiene/check-pr-hygiene/check-module-size (repo hygiene gates). verifier=ceiling — the agent's self-judgement never substitutes for it.

    Claude Code
  10. dev-toolkit1stars

    Слэш-команды code review / подготовки PR, скилл многоуровневого ревью диффа (уровни low…max; low — быстрый безагентный проход, medium и выше — параллельные углы поиска с adversarial-верификацией находок), скилл сквозного закрытия пункта TODO (реализация → ревью → PR → проверки → мёрж → релиз → проверка результата), скилл оптимизации и реконсиляции проектной документации (CLAUDE.md/README/память), защитный хук и звуковые уведомления для повседневной разработки на PHP, Go и JS

    Claude Code4 skills
  11. chat-handoff1stars

    Скилл для переноса контекста текущего диалога в новый чат одним готовым к вставке markdown-блоком

    Claude Code1 skill
  12. agent-orchestration0stars

    A feature-shipping pipeline (planner, coder, tester, reviewer) driven by /ship: plan with a finalize gate, then ship one service or fan out across many in parallel (phase-batched), configurable via an optional agent-orchestration.config.json registry, coordinating through a shared ./pipeline bus with a live local dashboard.

    Claude Code1 skill
  13. healthypress1stars

    Run a private WordPress.com site as a personal health record: creates a dedicated privacy-hardened site with a closed health taxonomy, then records each event as a private, dated, tagged post. Records and organizes; never diagnoses or advises.

    Claude Code3 skills1 MCP server
  14. cw1stars

    Claude Code workflow plugin: design/build/ship skills, orchestration agents, and guardrail hooks.

    Claude Code6 skills
  15. grace246stars

    Complete GRACE methodology: semantic markup, knowledge graphs, module contracts, testing, and log-driven verification for AI-driven engineering

    Claude Code
  16. verify-marketplace0stars

    Verify a plugin marketplace repo against the current Claude Code and Codex docs fetched live: catalog, manifests, layout, README, versions.

    Claude Code1 skill
  17. shortcuts0stars

    Short slash commands for the git, pull request, CI, docs, and dependency loop: /c, /cp, /cb, /cbp, /cpr, /cprw, /p, /m, /wm, /ci, /fci, /prd, /cl, /fa, /docs, /rule, /lint, /deps.

    Claude Code17 skills
  18. report-upstream0stars

    Take a bug found in a third-party dependency upstream: locate the repo, check for an existing report, reproduce, follow its contributing rules, draft the issue or PR for confirmation.

    Claude Code1 skill
  19. dependency-vetting0stars

    Verify a package or tool is authentic before installing, adding, or recommending it: follow the upstream project to its install command, never a registry page back to a project.

    Claude Code1 skill
  20. dependabot-review0stars

    Review Dependabot PRs for breaking changes, codebase impact, and merge readiness: one PR by URL, or an audit of every open one with a triage table.

    Claude Code1 skill
  21. cut-release0stars

    Cut a release from the Unreleased changelog section: version bump, release branch, PR, squash merge, tag, GitHub release.

    Claude Code1 skill
  22. browser-qa0stars

    Check a web page, or the pages this branch changed, in a real browser at mobile, tablet, and desktop widths, and report what is off.

    Claude Code1 skill
  23. browser-compat0stars

    Audit frontend code for 50 cross-browser bugs and mobile compatibility pitfalls.

    Claude Code1 skill
  24. atomic-commits0stars

    Commit early and often in atomic increments: separate refactor, feature, and cleanup work, keep PRs near 200 lines, and get a hook nudge when the diff grows past that.

    Claude Code1 skill
  25. codememory-plugin158stars

    CodeMemory plugin for Claude Code CLI. Replaces traditional sliding-window conversation compaction with a DAG-based summarization system that preserves every message while maintaining active context within model token limits.

    Claude Code5 skills
  26. ccxp-skills0stars

    Task lifecycle, PR automation, CI/RCA and retro skills for the ccxp-family workflow suite, plus general-purpose dev-tool skills (Cloudflare, docs, email, web-perf).

    Claude Code
  27. azure-cost252stars

    Azure cost analysis, estimation, optimization, and governance workflows backed by Azure Resource Manager MCP.

    Claude Code4 skills1 MCP server
  28. kingdom-factor0stars

    Kingdom Factor internal skill suite. kf-prospecting: operator console for the n8n prospect pipeline (what is stuck and why, fix a row, approve a person, start a run). kf-n8n-ops: keep the pipeline healthy — Tier 1 repairs the operator may do alone, Tier 2 changes go to Jordan. kf-marketing-manager: plain-English coach marketing reports via the kingdom-factor app's hosted, OAuth-protected, role-scoped remote MCP connector. kf-new-coach: onboard one coach across the GHL CRM, the Kingdom Factor platform and the Airtable Coaches index, as an audit the operator reads and then an apply she confirms.

    Claude Code4 skills2 MCP servers
  29. agent-router0stars

    Choose subagent models with /agent-models and click the prompt panel to inspect activity, token usage, and routing. Requires Claude Code Mod function hooks enabled.

    Claude Code
  30. hfi-skills0stars

    HFI's own Claude skills for investment work — from deal-pipeline tasks to standalone market analysis. Ask in plain English and the right skill runs, e.g. 'brief me on [deal] before my call' or 'build a trading-comps workbook for AAPL, MSFT, GOOGL.' Pipeline-aware skills pair with the hfi-deals plugin; others (like trading comps) work from primary sources such as SEC filings.

    Claude Code1 skill
  31. hfi-deals0stars

    Bring the HFI deal pipeline into ChatGPT and Codex: browse the board, search every deal and document, read OMs and meeting notes, and update deals in plain English. Sign in with your HFI Google account; changes attribute to you.

    CodexClaude Code1 MCP server
  32. aws-startups-solution-architecture19stars

    Technical AWS solutions for the problems startups get stuck on, from the AWS Startups Solution Architecture team. Multi-tenant SaaS isolation enforced in IAM rather than application code, and running a judgment agent such as an LLM reviewer inside a CI path. Startup here means the technical constraints are startup constraints: no dedicated platform or security team, a cost ceiling that is someone's runway, and a deadline that is a funding milestone. Draws all general-purpose AWS service depth from Agent Toolkit for AWS as an upstream dependency rather than restating it.

    Claude Code3 skills
  33. migration-to-aws19stars

    Migrate from GCP or Heroku to AWS — including your entire AI stack. Moves infrastructure (Cloud Run → Fargate, Heroku Dynos → Elastic Beanstalk by default with Fargate/EKS overrides, databases → RDS/Aurora), OpenAI/Gemini workloads to Amazon Bedrock, and agentic systems (LangChain, CrewAI, AutoGen, OpenAI Agents SDK) to AWS-native frameworks. Generates runnable Terraform, migration scripts, provider adapters, and deployment artifacts. Gives honest model-by-model pricing comparisons so you know exactly when Bedrock saves money and when it doesn't.

    CodexClaude Code2 skills4 MCP servers
  34. aws-startup-advisor19stars

    Personalized AWS guidance built on patterns from 350,000+ startups—architecture, cost, security, and migration, from day-one account setup to production-ready infrastructure. Migrate to AWS from GCP or Heroku—including your entire AI stack: OpenAI/Gemini/Anthropic SDK rewrites to Amazon Bedrock, agentic systems (LangChain, CrewAI, AutoGen) to AWS-native runtimes, and Temporal workers. Includes AWS Activate credits eligibility and 60+ exclusive startup offers. Built by AWS Startup Solutions Architects, with multi-account, multi-region support.

    CodexClaude Code7 skills4 MCP servers
  35. god-prompt1stars

    A scoped production-engineering workflow for implementing, debugging, reviewing, refactoring, and release-preparing software with explicit scope and verification evidence.

    CodexAgent Plugins1 skill
  36. god-prompt-mcp1stars

    GodPrompt engineering workflows for task routing, TDD, debugging, verification, and progressive disclosure through MCP.

    Claude CodeAgent Plugins1 skill1 MCP server
  37. rhapsody1stars

    Skills for Rhapsody Teams — how named agent teammates get work done (routing, the room, memory, the two review models), how to compose a roster and write a teammate profile, and how to operate a running installation.

    Claude Code2 skills
  38. wadb5stars

    Teach Claude Code to pair, reconnect, inspect, diagnose, and disconnect Android devices over ADB Wi-Fi with wadb.

    Claude Code1 skill
  39. groupsplit1stars

    Drive the GroupSplit CLI from an agent: the JSON contract, the exit codes, the confirmation protocol for destructive changes, and filing imported bank rows without recording the same money twice.

    Claude Code2 skills
  40. ultimate-seo-geo78stars

    Scored SEO audits for developers and marketers using AI agents: technical health, Schema.org JSON-LD, E-E-A-T, backlinks, domain signals, GEO (AI Overviews, ChatGPT, Perplexity). 56 bundled scripts; HTML/Excel/PDF reports. E-commerce schema, drift monitoring, topic clustering, Google API tiers (GSC/GA4), maps intelligence. Optional MCP (Firecrawl, DataForSEO): references/optional-extensions-mcp.md. Modes: Audit → Plan → Execute.

    Claude Code1 skill
  41. nautilus-compass207stars

    Cross-agent memory layer with drift detection, Merkle integrity, semantic recall, and MCP A2A protocol. Surfaces user-facing slash commands for direct inspection of session memory.

    Claude Code2 skills
  42. marketbase1stars

    Your entire GTM context, in one place. Bootstrap a MarketBase database, source and qualify leads with full provenance, and stage them into campaigns.

    Claude Code1 skill
  43. architecture-scan0stars

    Generic, resumable architecture/codebase assessment scan. Produces evidence-based findings and an upgrade roadmap for any codebase, in any technology.

    Claude Code1 skill
  44. omni38stars

    Full Omni platform control — three-tier skill system for agents (omni-agent), first-time setup (omni-setup), and platform ops (omni-ops)

    Claude Code4 skills
  45. advisor-decision0stars

    Пять линз (вероятностная, системная, процессная, выпуклость, когнитивные искажения) и три структурные роли разбирают решение независимо, скептики перекрёстно проверяют утверждения, валидатор с роутером Канемана — Кляйна взвешивает линзы и пишет один вердикт файлом.

    Claude Code
  46. jadlis-claudecode0stars

    Первый шаг маршрута Jadlis: собирает Claude Code, терминал и папку ~/Jadlis один в один с машиной владельца — зависимости, слияние ~/.claude/settings.json, правила, стиль ответа, хуки, строка состояния, tmux, динамический профиль iTerm2 и команда cld.

    Claude Code1 skill
  47. biomcp640stars

    Biomedical search and evidence retrieval MCP server

    Claude Code
  48. swd0stars

    Software development skills: product specs, deep planning, root-cause analysis, agent-oriented repo docs, spec-aware rebasing, PR review, PR submission, review revision, and end-user e2e testing.

    Agent Plugins1 skill
  49. with-agents0stars

    with-agents is an open-source Agent Plugins plugin indexed from GitHub with 3 skills. Review its manifest and source before installing.

    Agent Plugins3 skills
  50. valid-plugin0stars

    valid-plugin is an open-source Agent Plugins plugin indexed from GitHub with 1 skill. Review its manifest and source before installing.

    Agent Plugins1 skill