{"id":51286455,"url":"https://github.com/focus-mcp/marketplace","last_synced_at":"2026-06-30T06:00:48.593Z","repository":{"id":351382042,"uuid":"1210470024","full_name":"focus-mcp/marketplace","owner":"focus-mcp","description":"68+ atomic MCP bricks for AI agents — files, code intel, shell, reasoning, search, graphs, workflows. Official FocusMCP catalog.","archived":false,"fork":false,"pushed_at":"2026-06-26T09:28:45.000Z","size":2549,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-26T11:11:41.603Z","etag":null,"topics":["agent-tools","ai","bricks","catalog","claude","context-engineering","focus","llm","marketplace","mcp","model-context-protocol","npm-packages","plugins","tools","typescript"],"latest_commit_sha":null,"homepage":"https://focusmcp.dev","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/focus-mcp.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null,"governance":null,"roadmap":"ROADMAP.md","authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2026-04-14T12:56:55.000Z","updated_at":"2026-05-24T21:29:46.000Z","dependencies_parsed_at":null,"dependency_job_id":"3d701d8b-b6bd-40f7-8211-93273e89f4dc","html_url":"https://github.com/focus-mcp/marketplace","commit_stats":null,"previous_names":["focus-mcp/marketplace"],"tags_count":7,"template":false,"template_full_name":null,"purl":"pkg:github/focus-mcp/marketplace","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/focus-mcp%2Fmarketplace","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/focus-mcp%2Fmarketplace/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/focus-mcp%2Fmarketplace/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/focus-mcp%2Fmarketplace/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/focus-mcp","download_url":"https://codeload.github.com/focus-mcp/marketplace/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/focus-mcp%2Fmarketplace/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34915002,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-29T02:00:05.398Z","response_time":58,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-tools","ai","bricks","catalog","claude","context-engineering","focus","llm","marketplace","mcp","model-context-protocol","npm-packages","plugins","tools","typescript"],"created_at":"2026-06-30T06:00:47.833Z","updated_at":"2026-06-30T06:00:48.581Z","avatar_url":"https://github.com/focus-mcp.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003c!--\nSPDX-FileCopyrightText: 2026 FocusMCP contributors\nSPDX-License-Identifier: MIT\n--\u003e\n\n# FocusMCP Marketplace\n\n\u003e Official catalog of 68+ bricks for [FocusMCP](https://github.com/focus-mcp/cli).\n\n![Built with Claude Code](https://img.shields.io/badge/built_with-Claude_Code-8A2BE2)\n\n## What's a brick?\n\nA brick is an atomic MCP module with a well-defined scope.\nExamples: `shell` runs commands, `echo` replies messages, `lastversion` checks package versions.\n\nUsers compose bricks on demand — only the ones they need load into their AI's context.\n\n## Browse bricks\n\nUse the FocusMCP CLI:\n\n```bash\nnpm install -g @focus-mcp/cli\nfocus browse\n```\n\nOr fetch the catalog directly:\n\n```bash\ncurl https://focus-mcp.github.io/marketplace/catalog.json\n```\n\n## Categories\n\n- **Files**: fileread, filewrite, filelist, fileops, filesearch, filediff, smartread, multiread\n- **Code Intelligence**: treesitter, symbol, callgraph, depgraph, cache, outline, refs\n- **Context**: smartcontext, overview, compress, tokenbudget, contextpack\n- **Shell \u0026 Execution**: shell, sandbox, batch\n- **Code Editing**: rename, codeedit, inline\n- **Reasoning**: thinking, planning, decision\n- **Search**: textsearch, fts, semanticsearch\n- **Knowledge**: knowledge, graphbuild, graphquery, graphcluster, graphexport\n- **Orchestration**: dispatch, parallel, debate, review, research, agent, share, task\n- **Analytics**: metrics, heatmap, savings\n- **Utilities**: format, validate, convert, diagram, routes\n- **Workflows**: onboarding, fullaudit, autopilot\n- **Version intelligence**: lastversion\n\n(This list is for humans — the full catalog is at `catalog.json`)\n\n## Measured economy\n\nStatic, reproducible benchmark on 79 tools across all 69 bricks (no LLM, no variance):\n\n| | Tokens |\n|---|---:|\n| Native equivalent (raw `Read`, `Grep`, etc.) | 366,912 |\n| FocusMCP bricks output | 127,228 |\n| **Saved** | **239,684 (−65.3%)** |\n\n**Top 5 most efficient tools** (best replacement for native operations):\n\n| Tool | Native | Brick | Δ |\n|---|---:|---:|---:|\n| `treesitter.ts_reindex` | 8,522 | 4 | **−100.0%** |\n| `treesitter.ts_index` | 5,645 | 7 | **−99.9%** |\n| `symbol.sym_find` | 2,678 | 4 | **−99.9%** |\n| `routes.rt_list` | 6,796 | 10 | **−99.9%** |\n| `symbol.sym_bulk` | 8,237 | 12 | **−99.9%** |\n\n**Caveats** (full transparency):\n- 4 tools show marginal overhead (+2 to +28%) on tiny outputs where JSON envelope weighs more than the operation. Examples: `fl_list`, `fr_head`, `fr_range`, `fmt_json`.\n- This is iso-call measurement: it captures payload economy, not multi-turn agent dynamics. Real agent sessions have variance and exploration costs that this benchmark doesn't model.\n- See [`benchmarks/equivalence-report.md`](./benchmarks/equivalence-report.md) for the full per-tool table and methodology.\n\nReproduce: `pnpm --filter @focusmcp/bench-harness exec tsx src/equivalence.ts`\n\n## Break-even by tool category\n\nSome bricks save tokens immediately, others need multiple calls to amortize their fixed cost\n(manifest loaded once per session = fixed overhead M+D).\n\n| Category | Tools | Break-even | Example |\n|---|---|---|---|\n| Always wins | outline, smartread, refs, compress, fts, textsearch, filesearch, overview, rename | N=1 | `outline.out_repo` saves 23,226 tokens/call |\n| Wins fast (≤10 calls) | — | ≤ 10 | — |\n| Wins eventually (≤100 calls) | — | ≤ 100 | — |\n| Never wins (single-shot) | fileread, filelist, multiread, filediff, format | ∞ | JSON envelope \u003e operation savings |\n\n**Session projection** (50 varied calls, top-quartile mix): ~80% token savings vs native.\n\nSee [full math model report](./benchmarks/equivalence-math-report.md).\n\n## Roadmap \u0026 improvements\n\nEnhancement opportunities tracked in [`IMPROVEMENTS.md`](./IMPROVEMENTS.md) — architecture, performance, and new capabilities.\nBugs and patches are in [`benchmarks/PATCH_QUEUE.md`](./benchmarks/PATCH_QUEUE.md).\n\n\n## Contribute a brick\n\nSee [CONTRIBUTING.md](./CONTRIBUTING.md).\n\n## AI-assisted development\n\nFocusMCP was built with heavy Claude Code assistance — its architecture, implementation,\ndocs, and tests have all been co-authored with AI. We embrace this openly because:\n\n1. **Transparency matters** — we'd rather disclose it than pretend otherwise\n2. **AI tooling is the context** — we're building tools for AI agents, it makes sense to use them\n3. **Quality over origin** — what matters is that the code is tested, reviewed, and working\n\n**Your AI-assisted contributions are welcome.** We don't require you to hide the fact that\nClaude, Copilot, Cursor, or any other tool helped you. What we do expect:\n\n- Tests pass, code is typed, lint is green\n- You've read the diff and understand what the PR does\n- Conventional Commits, clear PR description\n- You can explain your design choices during review\n\nSee [CONTRIBUTING.md](./CONTRIBUTING.md) for the full guidelines.\n\n## License\n\n[MIT](./LICENSE)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffocus-mcp%2Fmarketplace","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ffocus-mcp%2Fmarketplace","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffocus-mcp%2Fmarketplace/lists"}