agent-vitals
Provides daily census and time-series statistics on GitHub repositories in the AI agent tooling ecosystem, including maintenance status, licence analysis, and repository churn.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
agent-vitals
A daily census of the AI agent tooling ecosystem on GitHub: MCP servers, agent frameworks, skills and the tools around them. It measures which of them are still alive.
Every awesome-list tells you what was published. None of them tell you what has been touched since. This repository answers three questions every day, from public GitHub metadata, and keeps the answers as a time series:
How much of the ecosystem is still maintained?
How much of it carries a licence you could actually use at work?
What appeared, and what went quiet, since yesterday?
2026-11-03
50,701 repositories across 13 topic queries in 2 tiers: mcp (23,425, 2+ stars), agents (27,276, 10+ stars).
48.8% pushed in the last 30 days.
7,546 (14.9%) have no licence file at all, which leaves them under exclusive copyright by default: no permission to use, copy or modify them, whatever the README suggests. A further 4,347 (8.6%) carry a licence GitHub cannot map to a standard identifier. Those are licensed, just not in a way a procurement review waves through.
Only 9,738 of these repositories are even a year old. Among those, 28.1% have not been pushed since. The headline 5.4% across the whole index is an artefact of how young this ecosystem is.
Churn since the previous run: +490 new, -64 gone.
Maintenance status
Status | Repositories | Share | Meaning |
active | 24,737 | 48.8% | pushed within 30 days |
slowing | 9,362 | 18.5% | 31-90 days |
stale | 13,035 | 25.7% | 91-365 days |
abandoned | 2,734 | 5.4% | no push in over a year |
archived | 833 | 1.6% | archived by its owner |
Licences
no licence file and licence GitHub cannot identify are different things and are
counted separately. The second group has a LICENSE file; GitHub simply cannot match
it to a standard identifier.
Licence | Repositories | Share |
MIT | 28,403 | 56.0% |
no licence file | 7,546 | 14.9% |
Apache-2.0 | 7,168 | 14.1% |
licence GitHub cannot identify | 4,347 | 8.6% |
AGPL-3.0 | 1,443 | 2.8% |
GPL-3.0 | 804 | 1.6% |
CC0-1.0 | 216 | 0.4% |
BSD-3-Clause | 138 | 0.3% |
MPL-2.0 | 118 | 0.2% |
MIT-0 | 94 | 0.2% |
Most-starred, still maintained
Ranked by stars, restricted to repositories pushed within the last 30 days.
Repository | Stars | Licence | Last push | Description |
425,180 | MIT | 2026-11-03 | Reverse engineer anything with agents, from app behavior down to native binaries. | |
296,113 | MIT | 2026-11-02 | The agent harness performance optimization system. Skills, instincts, memory, security, and research-first dev | |
270,552 | MIT | 2026-11-03 | DeepSeek Harness: Everything is a Plugin. | |
260,071 | MIT | 2026-11-03 | The agent that grows with you | |
209,431 | non-standard | 2026-11-03 | Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, | |
198,173 | AGPL-3.0 | 2026-11-03 | Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥 | |
187,885 | non-standard | 2026-11-03 | AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the too | |
186,971 | no licence file | 2026-10-28 | Public repository for Agent Skills | |
183,329 | MIT | 2026-10-29 | Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote. | |
160,317 | non-standard | 2026-11-03 | Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. De | |
159,402 | Apache-2.0 | 2026-10-25 | Java 面试 & 后端通用面试指南,覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发 | |
157,635 | non-standard | 2026-11-03 | User-friendly AI Interface (Supports Ollama, OpenAI API, ...) | |
151,448 | MIT | 2026-11-03 | A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode,DeepSeek Harness and more. Only | |
148,937 | MIT | 2026-11-03 | The agent engineering platform. | |
144,562 | MIT | 2026-11-02 | An AI skill that provides design intelligence for building professional UI/UX across multiple platforms. | |
141,544 | Apache-2.0 | 2026-11-02 | Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphif | |
119,921 | MIT | 2026-11-02 | Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent | |
117,079 | Apache-2.0 | 2026-11-03 | 🪨 why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by | |
114,080 | MIT | 2026-11-03 | Orca is the place for working with a fleet of agents (e.g. Grok Bot, Claude Code). Run any coding agent with y | |
112,936 | MIT | 2026-11-01 | Production-grade engineering skills for AI coding agents. | |
111,332 | Apache-2.0 | 2026-11-03 | Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, comp | |
110,702 | MIT | 2026-11-03 | The open-source app everyone uses to manage agents at work | |
109,204 | MIT | 2026-11-03 | Agent skills that give coding agents taste: art direction, reference images and a browser check instead of gen | |
108,700 | Apache-2.0 | 2026-11-03 | An open-source AI agent that brings the power of Gemini directly into your terminal. | |
108,271 | MIT | 2026-11-01 | Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, |
Popular but unmaintained
Over 50 stars, no push in more than a year. These are the entries that stay on curated lists long after anyone stopped answering issues.
Repository | Stars | Last push | Days |
17,729 | 2025-01-22 | 650 | |
15,893 | 2025-10-03 | 396 | |
8,101 | 2025-02-10 | 631 | |
6,796 | 2025-07-13 | 478 | |
5,968 | 2024-09-26 | 768 | |
5,016 | 2025-10-13 | 386 | |
4,394 | 2025-10-26 | 373 | |
4,070 | 2025-10-28 | 371 | |
3,579 | 2025-04-23 | 559 | |
3,576 | 2025-10-09 | 390 | |
3,502 | 2025-07-27 | 464 | |
2,598 | 2024-11-07 | 726 | |
2,473 | 2024-04-29 | 918 | |
2,350 | 2024-05-01 | 916 | |
2,308 | 2024-12-11 | 692 | |
2,149 | 2025-08-21 | 439 | |
2,111 | 2025-04-22 | 560 | |
1,901 | 2024-07-10 | 846 | |
1,744 | 2025-10-20 | 379 | |
1,705 | 2025-08-15 | 445 |
Quiet, permissive, still asked about
No push in over a year, or archived, yet 200 or more stars, a permissive licence and five or more open issues. People still arrive; nobody answers. Each is forkable as it stands. The list reads a date, a star count and a licence field; it does not judge the work.
Repository | Stars | Open issues | Licence | Last push |
17,729 | 270 | MIT | 2025-01-22 | |
15,893 | 69 | MIT | 2025-10-03 | |
11,068 | 73 | Apache-2.0 | 2025-09-12 | |
8,101 | 86 | MIT | 2025-02-10 | |
6,796 | 268 | MIT | 2025-07-13 | |
6,545 | 103 | MIT | 2026-06-05 | |
6,162 | 53 | MIT | 2026-03-23 | |
5,968 | 70 | Apache-2.0 | 2024-09-26 | |
4,394 | 59 | MIT | 2025-10-26 | |
4,098 | 85 | MIT | 2026-08-06 | |
4,070 | 149 | Apache-2.0 | 2025-10-28 | |
3,698 | 16 | Apache-2.0 | 2026-10-24 | |
3,398 | 52 | Apache-2.0 | 2026-07-20 | |
2,598 | 22 | MIT | 2024-11-07 | |
2,438 | 6 | Apache-2.0 | 2025-06-09 | |
2,350 | 23 | Apache-2.0 | 2024-05-01 | |
2,242 | 43 | MIT | 2025-04-15 | |
2,149 | 48 | MIT | 2025-08-21 | |
1,901 | 7 | MIT | 2024-07-10 | |
1,851 | 30 | Apache-2.0 | 2026-08-13 |
Busy, permissive, with open work
Pushed in the last 30 days, 1,000 or more stars, a permissive licence and thirty or more open issues: maintained, wanted, and short of hands. Ranked by open issues.
Repository | Open issues | Stars | Licence | Last push |
52,486 | 260,071 | MIT | 2026-11-03 | |
7,996 | 114,080 | MIT | 2026-11-03 | |
6,637 | 110,702 | MIT | 2026-11-03 | |
5,409 | 13,291 | Apache-2.0 | 2026-11-03 | |
4,535 | 37,626 | MIT | 2026-11-03 | |
4,504 | 2,115 | Apache-2.0 | 2026-10-25 | |
4,121 | 97,239 | MIT | 2026-11-02 | |
3,172 | 4,749 | Apache-2.0 | 2026-11-03 | |
2,920 | 28,896 | MIT | 2026-11-03 | |
2,718 | 31,877 | MIT | 2026-10-28 | |
2,612 | 2,427 | Apache-2.0 | 2026-10-25 | |
2,529 | 11,184 | Apache-2.0 | 2026-11-03 | |
2,528 | 9,104 | Apache-2.0 | 2026-11-03 | |
2,099 | 28,932 | Apache-2.0 | 2026-11-03 | |
1,967 | 43,031 | Apache-2.0 | 2026-11-03 | |
1,798 | 21,006 | MIT | 2026-11-03 | |
1,773 | 2,499 | MIT | 2026-11-02 | |
1,701 | 1,898 | MIT | 2026-11-03 | |
1,679 | 425,180 | MIT | 2026-11-03 | |
1,668 | 85,410 | Apache-2.0 | 2026-11-03 |
Trend
Date | Repositories | Active | Abandoned | No licence file | New |
2026-10-21 | 46,529 | 22,246 | 2,636 | 7,134 | +449 |
2026-10-22 | 46,834 | 22,508 | 2,643 | 7,146 | +392 |
2026-10-23 | 47,119 | 22,725 | 2,647 | 7,178 | +343 |
2026-10-24 | 47,363 | 22,863 | 2,654 | 7,206 | +292 |
2026-10-25 | 47,671 | 23,062 | 2,658 | 7,294 | +348 |
2026-10-26 | 46,953 | 22,851 | 2,603 | 7,199 | +298 |
2026-10-27 | 47,223 | 23,094 | 2,514 | 7,175 | +1263 |
2026-10-28 | 48,474 | 23,531 | 2,673 | 7,365 | +1328 |
2026-10-29 | 48,627 | 23,499 | 2,659 | 7,334 | +360 |
2026-10-30 | 49,315 | 23,910 | 2,695 | 7,436 | +751 |
2026-10-31 | 49,711 | 24,072 | 2,709 | 7,511 | +453 |
2026-11-01 | 49,976 | 24,184 | 2,711 | 7,524 | +326 |
2026-11-02 | 50,275 | 24,400 | 2,720 | 7,530 | +370 |
2026-11-03 | 50,701 | 24,737 | 2,734 | 7,546 | +490 |
Related MCP server: @aipulsegeorgia/mcp-server
Ask it from an agent
The census is an MCP server too, standard library only, four tools: lookup a
repository's status and licence, search by words with status and licence
filters, summary of the latest run, and the two candidates lists.
git clone https://github.com/Keremozdemirra/agent-vitals && cd agent-vitals
claude mcp add agent-vitals -- python3 "$PWD/mcp_server.py"AGENT_VITALS_REMOTE=1 reads the published index instead of the checkout, so
the clone can stay old. Nothing is written or executed; it answers from
data/servers.json.
Use the data
curl -sL https://raw.githubusercontent.com/Keremozdemirra/agent-vitals/main/data/servers.csv -o servers.csvFile | What it is |
| Full index, one object per repository, with |
| The same index, flat, for spreadsheets and |
| One row per day: totals per status, licence counts, churn. |
| That day's snapshot, including which repositories arrived and which went quiet. |
| The two acted-on lists from the report: |
Everything is committed, so git log data/history.csv is the changelog of the
ecosystem itself.
Method
Source: the GitHub REST search API, public repository metadata only. Nothing is cloned, downloaded or executed.
Tiers and their star floors:
mcp(2+ stars):topic:mcp-server,topic:model-context-protocolagents(10+ stars):topic:mcp,topic:claude-code,topic:ai-agents,topic:ai-agent,topic:agent-skills,topic:agentic-ai,topic:claude-skills,topic:autonomous-agents,topic:llm-agents,topic:agent-framework,topic:llm-toolsThe MCP topics name one specific thing, so two stars is enough. The broad agent topics are also attached to every tutorial and course repository in the field; ten stars is where they start describing tools rather than exercises.
Each query is sliced by star count until every slice fits under GitHub's 1000-result ceiling, so this is a census rather than a top-1000 sample.
statusis derived frompushed_atalone. It measures whether a repository is being touched, not whether it is good. A finished, correct tool can sit atabandonedand still work.license_stateis the field that matters.spdxmeans GitHub matched a standard licence;non-standardmeans a LICENSE file exists that GitHub reports asNOASSERTION(n8n's Sustainable Use License, for one);nonemeans no licence file was detected at all. Only the last of those means "you have no permission to use this", and conflating the two would misrepresent projects that did license their work.A repository leaves the index when it is deleted, renamed, drops below the star floor, or has its topic removed.
left_the_indexdoes not meandeleted, and this index cannot tell those cases apart.Repository descriptions are third-party text. They are stripped of control characters, truncated, and escaped at render time.
What this is not
Not a security audit. Nothing here says a repository is safe or unsafe. The fields are dates, counts and licence identifiers: facts from the API. None of them judges anyone's code.
Not a quality ranking. Stars measure attention, not merit.
Not a recommendation. Check anything you install yourself.
What is in here, and whose it is
Nothing in this repository is anyone else's work. No repository is cloned, downloaded, or copied. Two kinds of thing are published:
Facts from the GitHub API: name, URL, star and fork counts, creation and last-push dates, detected licence identifier, archived flag. Facts about public repositories, not expression, and each one links to its source.
The repository's own one-line description, as its author wrote it, truncated and shown next to a link to the original. This is the only third-party text here, and it is used the way every package registry and search index uses it.
Everything else (the collector, the renderer, this text, the analysis) was written for this repository.
Licence
Code (
collect.py,render.py, the workflow): MIT.The compilation (the selection, structure and derived fields in
data/): CC0 1.0. Take it, chart it, fork it, no attribution required.The
descriptionfield: belongs to whoever wrote it, and is reproduced here as a short factual descriptor alongside a link to its source. It is not covered by the CC0 grant above, and this index makes no claim over it.The indexed repositories themselves: their authors', under their own licences, which is precisely what this index measures.
If you own a repository listed here and want its description dropped from the index, open an issue and it will be removed from the next run.
Regenerated automatically. Last run: 2026-11-03T12:35:13+00:00 · status: partial.
This server cannot be deployed
Maintenance
Related MCP Connectors
Agent payments ecosystem intelligence. Scans GitHub/HN/npm across AP2, ACP, x402, MPP, UCP.
Daily AI-ecosystem data: GitHub AI/MCP repos, MCP registry, HF trending, LLM pricing, with history.
Search and fetch AI agent skills, rules files and MCP servers indexed from GitHub.
Search agent-ready open source repos by task, interface, ecosystem and OG score.
Related MCP Servers
AlicenseAqualityCmaintenanceEnables coding agents to search an index of 22k+ MCP servers, agent skills and coding-agent plugins directly from chat, returning matches with stars, install commands and page links, plus trending repos and curated best-of topic lists. Read-only, no API key required, usable via a hosted HTTP endpoint or locally over stdio.4MIT- FlicenseNot gradedqualityAmaintenanceEnables natural-language search of 378 curated AI agent frameworks, developer tools, and automation repositories from inside Claude Code, Cursor, Codex, or any MCP-compatible client, returning matching repos with descriptions instead of manual GitHub browsing. Also exposes category listing, stats, and single-repo lookup.-
- FlicenseNot gradedqualityAmaintenanceEnables searching, listing, filtering, and retrieving details of 378 curated AI agent frameworks, developer tools, and automation resources directly from Claude Code, Cursor, or any MCP-compatible client using natural-language queries. It returns bilingual (English/Georgian) descriptions and repository statistics so assistants can recommend the right repo without leaving the conversation.-
- FlicenseNot gradedqualityAmaintenanceEnables Claude Code, Cursor, Codex, and any MCP-compatible client to search, list, and retrieve entries from a curated collection of 378 AI agent frameworks, developer tools, and automation repositories using natural language, with category filters and English/Georgian descriptions. Also exposes stats and per-repo details so assistants can answer questions like which resource is best for RAG or memory.-