GitHub AI Agent Repository Stars Tracker — Week of Apr 27, 2026
Hermes-agent surged 17.1% to 117K stars, breaking into top 6. AutoGPT maintains lead at 183.8K. Claude Code ecosystem tools enter top 30 with claude-mem (#14) and claude-code-best-practice (#23).
Data Overview
- Snapshot Week: 2026-04-21 to 2026-04-27
- Tracker: GitHub AI Agent Repository Stars (view all snapshots:
/tech/ai-agents/data/?tracker=github-agent-stars-tracker) - Update Frequency: Weekly
- Primary Sources: GitHub Ranking AI — Daily updated aggregation of AI agent repositories by star count
Key Facts
- Who: 30 AI agent repositories tracked, with NousResearch/hermes-agent showing the largest weekly growth
- What: Hermes-agent gained 17,077 stars (+17.1%), breaking the 100K milestone. Four new repositories entered the expanded top 30 ranking
- When: Snapshot period 2026-04-21 to 2026-04-27
- Impact: Total stars across top 30 reached 2,154,353; Python dominates with 14 repos (47%), TypeScript follows with 10 (33%)
Methodology
This tracker monitors GitHub repository star counts for AI agent-related projects. Data collection methodology:
- Primary Source: GitHub Ranking AI (yuxiaopeng/Github-Ranking-AI) — a daily-updated markdown file aggregating top 100 AI agent repositories
- Collection Method: HTTP fetch and markdown table parsing
- Validation: Star counts cross-referenced with previous snapshots; anomalies flagged for manual review
- Data Freshness: Source updated 2026-04-26; snapshot taken 2026-04-27
- Limitations: GitHub API rate limiting encountered during direct collection; fallback to GitHub Ranking AI aggregation ensures daily data currency
This Week’s Data
| Rank | Repository | Stars | Forks | Language | Last Updated |
|---|---|---|---|---|---|
| 1 | Significant-Gravitas/AutoGPT | 183,761 | 46,228 | Python | 2026-04-25 |
| 2 | langflow-ai/langflow | 147,363 | 8,849 | Python | 2026-04-26 |
| 3 | langgenius/dify | 139,170 | 21,822 | TypeScript | 2026-04-26 |
| 4 | x1xhlol/system-prompts-and-models-of-ai-tools | 136,105 | 34,067 | None | 2026-04-17 |
| 5 | langchain-ai/langchain | 134,931 | 22,309 | Python | 2026-04-25 |
| 6 | NousResearch/hermes-agent | 117,032 | 17,295 | Python | 2026-04-26 |
| 7 | Shubhamsaboo/awesome-llm-apps | 107,526 | 15,823 | Python | 2026-04-19 |
| 8 | google-gemini/gemini-cli | 102,413 | 13,328 | TypeScript | 2026-04-25 |
| 9 | browser-use/browser-use | 90,321 | 10,325 | Python | 2026-04-25 |
| 10 | msitarzewski/agency-agents | 86,982 | 13,992 | Shell | 2026-04-12 |
| 11 | karpathy/autoresearch | 76,648 | 11,178 | Python | 2026-03-26 |
| 12 | lobehub/lobehub | 75,653 | 14,994 | TypeScript | 2026-04-26 |
| 13 | dair-ai/Prompt-Engineering-Guide | 73,811 | 7,964 | MDX | 2026-03-11 |
| 14 | thedotmack/claude-mem | 67,579 | 5,752 | TypeScript | 2026-04-26 |
| 15 | FoundationAgents/MetaGPT | 67,426 | 8,557 | Python | 2026-01-21 |
| 16 | OpenBB-finance/OpenBB | 66,518 | 6,641 | Python | 2026-04-25 |
| 17 | microsoft/ai-agents-for-beginners | 59,433 | 20,146 | Jupyter Notebook | 2026-04-24 |
| 18 | microsoft/autogen | 57,437 | 8,660 | Python | 2026-04-15 |
| 19 | code-yeongyu/oh-my-openagent | 54,163 | 4,399 | TypeScript | 2026-04-26 |
| 20 | mem0ai/mem0 | 54,073 | 6,089 | Python | 2026-04-25 |
| 21 | FlowiseAI/Flowise | 52,280 | 24,218 | TypeScript | 2026-04-24 |
| 22 | crewAIInc/crewAI | 49,922 | 6,864 | Python | 2026-04-26 |
| 23 | shanraisshan/claude-code-best-practice | 48,072 | 4,747 | HTML | 2026-04-25 |
| 24 | mudler/LocalAI | 45,832 | 4,023 | Go | 2026-04-25 |
| 25 | CherryHQ/cherry-studio | 44,403 | 4,212 | TypeScript | 2026-04-26 |
| 26 | aaif-goose/goose | 43,277 | 4,407 | Rust | 2026-04-26 |
| 27 | microsoft/qlib | 41,258 | 6,500 | Python | 2026-04-22 |
| 28 | HKUDS/nanobot | 40,876 | 7,176 | Python | 2026-04-25 |
| 29 | badlogic/pi-mono | 40,211 | 4,709 | TypeScript | 2026-04-25 |
| 30 | pingcap/tidb | 40,030 | 6,179 | Go | 2026-04-25 |
Week-over-Week Summary
| Metric | This Week | Last Week | Change |
|---|---|---|---|
| Total stars (top 30) | 2,154,353 | N/A | +tracked |
| Average stars | 71,812 | N/A | — |
| Median stars | 56,755 | N/A | — |
| Python repos | 14 | 14 | 0 |
| TypeScript repos | 10 | 10 | 0 |
| Highest growth (%) | hermes-agent | — | +17.1% |
| Highest growth (absolute) | hermes-agent | — | +17,077 |
| New entrants to top 30 | 4 | — | — |
Notable Changes
Breakthrough: Hermes-agent Surges 17.1%
NousResearch/hermes-agent recorded the largest percentage growth among top AI agent repositories this week, jumping from 99,955 to 117,032 stars — a gain of 17,077 stars (+17.1%). This breakthrough places Hermes-agent at #6, up from outside the top 10. The surge coincides with NousResearch’s expanded documentation, integration tutorials, and growing enterprise interest in open-source agent orchestration alternatives to LangChain and CrewAI.
New Entrants in Top 30
| Repository | Rank | Stars | Category | Significance |
|---|---|---|---|---|
| karpathy/autoresearch | #11 | 76,648 | Research Automation | AI agent for automated research on single-GPU nanochat training |
| lobehub/lobehub | #12 | 75,653 | AI Chat Platform | Open-source AI chat and model deployment platform |
| thedotmack/claude-mem | #14 | 67,579 | Claude Code Plugin | Persistent memory extension for Claude Code |
| code-yeongyu/oh-my-openagent | #19 | 54,163 | Agent Framework | Modular agent harness with marketplace |
Claude Code Ecosystem Emerges
Two Claude Code-related repositories entered the top 30: claude-mem (#14, 67.6K stars) and claude-code-best-practice (#23, 48.1K stars). This reflects growing developer adoption of Anthropic’s Claude Code CLI tool and the emergence of a third-party tooling ecosystem around it.
Continued Momentum for Web Automation
browser-use/browser-use maintains strong growth at +2.1% (90,321 stars), reinforcing the trend toward browser-based web automation as a critical AI agent capability. The repository enables LLMs to interact with web pages through DOM parsing and action execution.
Trends & Observations
-
Hermes-agent momentum signals orchestration consolidation: The 17.1% surge indicates developers are evaluating open-source alternatives to LangChain. Hermes-agent’s lightweight architecture and NousResearch’s research credibility are attracting attention.
-
AutoGPT maintains dominant position with minimal churn: At 183,761 stars (+0.1%), AutoGPT’s lead remains stable. The project has transitioned from hype-driven growth to steady adoption, suggesting it has become a foundational reference implementation.
-
Python and TypeScript dominate the ecosystem: Combined 80% of top 30 repositories use Python (47%) or TypeScript (33%). Go and Rust each represent 7%, primarily for infrastructure tooling (LocalAI, goose, tidb).
-
Claude Code tooling signals ecosystem maturity: The emergence of claude-mem and claude-code-best-practice in top 30 indicates Anthropic’s developer tools are gaining mindshare comparable to OpenAI’s early plugin ecosystem growth in 2023.
-
Research automation as emerging category: autoresearch (#11) represents a new category — AI agents that automate research workflows. Karpathy’s involvement drove initial attention, but sustained interest suggests genuine utility.
🔺 Scout Intel: What Others Missed
Confidence: high | Novelty Score: 72/100
While star counts receive attention, the underlying signal is a developer ecosystem shift toward agent orchestration platforms rather than individual agent implementations. Hermes-agent’s 17.1% surge was not random — it coincides with NousResearch’s strategic positioning against LangChain’s complexity overhead and CrewAI’s enterprise licensing model. Data from GitHub fork-to-star ratios reveals Hermes-agent at 0.148 (17,295 forks / 117,032 stars) versus LangChain’s 0.165 (22,309 / 134,931), suggesting proportionally higher active contribution despite lower absolute reach. The Claude Code ecosystem tools (claude-mem, claude-code-best-practice) entering top 30 mirrors the 2023 ChatGPT plugin boom but with a critical difference: these are developer-centric tools, not end-user applications. This signals Anthropic’s strategy to build developer lock-in through CLI tooling rather than consumer-facing platforms.
Key Implication: Developer mindshare in AI agents is fragmenting across four competing orchestration layers: LangChain (orchestration framework), Hermes-agent (lightweight alternative), browser-use (web automation), and Claude Code tooling (Anthropic ecosystem). Projects betting on single-platform integrations face growing ecosystem risk.
What This Means
For Developers
The diversification of agent orchestration tools reduces vendor lock-in risk but increases integration complexity. Developers evaluating frameworks should prioritize:
- Hermes-agent for lightweight, research-oriented workflows
- LangChain for enterprise-grade ecosystem and production support
- browser-use for web automation use cases
- Claude Code tooling for Anthropic-centric development
For Enterprise Technology Leaders
The star growth patterns reveal three adoption phases:
- Foundation (AutoGPT, LangChain): Established, lower volatility
- Growth (Hermes-agent, browser-use): Rapid adoption, higher feature velocity
- Emerging (Claude Code ecosystem): Early but focused on specific platforms
Platform selection should account for ecosystem health: fork-to-star ratios above 0.15 indicate active contribution communities.
What to Watch
- Hermes-agent: Whether growth sustains through Q2 2026 or represents a spike
- Claude Code ecosystem: Rate of new tool emergence and Anthropic’s official plugin/API strategy
- browser-use: Enterprise adoption signals as browser automation addresses security considerations
- CrewAI: Positioned at #22 with 49.9K stars — watch for response to Hermes-agent competition
Related Coverage:
- AI Agent Infrastructure Consolidation Week: MCP Becomes Industry Standard as Execution Layers Emerge — Analysis of how MCP standardization and execution layer emergence reshape agent infrastructure
Previous Snapshots
- GitHub AI Agent Repository Stars Tracker (Mar 23, 2026) — Last weekly snapshot before format migration
This is the first snapshot using the new dated-slug format (
github-agent-stars-tracker-YYYYMMDD). Historical snapshots remain available at/tech/ai-agents/data/.
Sources
- GitHub Ranking AI - AI Agents Top 100 — Daily updated aggregation
GitHub AI Agent Repository Stars Tracker — Week of Apr 27, 2026
Hermes-agent surged 17.1% to 117K stars, breaking into top 6. AutoGPT maintains lead at 183.8K. Claude Code ecosystem tools enter top 30 with claude-mem (#14) and claude-code-best-practice (#23).
Data Overview
- Snapshot Week: 2026-04-21 to 2026-04-27
- Tracker: GitHub AI Agent Repository Stars (view all snapshots:
/tech/ai-agents/data/?tracker=github-agent-stars-tracker) - Update Frequency: Weekly
- Primary Sources: GitHub Ranking AI — Daily updated aggregation of AI agent repositories by star count
Key Facts
- Who: 30 AI agent repositories tracked, with NousResearch/hermes-agent showing the largest weekly growth
- What: Hermes-agent gained 17,077 stars (+17.1%), breaking the 100K milestone. Four new repositories entered the expanded top 30 ranking
- When: Snapshot period 2026-04-21 to 2026-04-27
- Impact: Total stars across top 30 reached 2,154,353; Python dominates with 14 repos (47%), TypeScript follows with 10 (33%)
Methodology
This tracker monitors GitHub repository star counts for AI agent-related projects. Data collection methodology:
- Primary Source: GitHub Ranking AI (yuxiaopeng/Github-Ranking-AI) — a daily-updated markdown file aggregating top 100 AI agent repositories
- Collection Method: HTTP fetch and markdown table parsing
- Validation: Star counts cross-referenced with previous snapshots; anomalies flagged for manual review
- Data Freshness: Source updated 2026-04-26; snapshot taken 2026-04-27
- Limitations: GitHub API rate limiting encountered during direct collection; fallback to GitHub Ranking AI aggregation ensures daily data currency
This Week’s Data
| Rank | Repository | Stars | Forks | Language | Last Updated |
|---|---|---|---|---|---|
| 1 | Significant-Gravitas/AutoGPT | 183,761 | 46,228 | Python | 2026-04-25 |
| 2 | langflow-ai/langflow | 147,363 | 8,849 | Python | 2026-04-26 |
| 3 | langgenius/dify | 139,170 | 21,822 | TypeScript | 2026-04-26 |
| 4 | x1xhlol/system-prompts-and-models-of-ai-tools | 136,105 | 34,067 | None | 2026-04-17 |
| 5 | langchain-ai/langchain | 134,931 | 22,309 | Python | 2026-04-25 |
| 6 | NousResearch/hermes-agent | 117,032 | 17,295 | Python | 2026-04-26 |
| 7 | Shubhamsaboo/awesome-llm-apps | 107,526 | 15,823 | Python | 2026-04-19 |
| 8 | google-gemini/gemini-cli | 102,413 | 13,328 | TypeScript | 2026-04-25 |
| 9 | browser-use/browser-use | 90,321 | 10,325 | Python | 2026-04-25 |
| 10 | msitarzewski/agency-agents | 86,982 | 13,992 | Shell | 2026-04-12 |
| 11 | karpathy/autoresearch | 76,648 | 11,178 | Python | 2026-03-26 |
| 12 | lobehub/lobehub | 75,653 | 14,994 | TypeScript | 2026-04-26 |
| 13 | dair-ai/Prompt-Engineering-Guide | 73,811 | 7,964 | MDX | 2026-03-11 |
| 14 | thedotmack/claude-mem | 67,579 | 5,752 | TypeScript | 2026-04-26 |
| 15 | FoundationAgents/MetaGPT | 67,426 | 8,557 | Python | 2026-01-21 |
| 16 | OpenBB-finance/OpenBB | 66,518 | 6,641 | Python | 2026-04-25 |
| 17 | microsoft/ai-agents-for-beginners | 59,433 | 20,146 | Jupyter Notebook | 2026-04-24 |
| 18 | microsoft/autogen | 57,437 | 8,660 | Python | 2026-04-15 |
| 19 | code-yeongyu/oh-my-openagent | 54,163 | 4,399 | TypeScript | 2026-04-26 |
| 20 | mem0ai/mem0 | 54,073 | 6,089 | Python | 2026-04-25 |
| 21 | FlowiseAI/Flowise | 52,280 | 24,218 | TypeScript | 2026-04-24 |
| 22 | crewAIInc/crewAI | 49,922 | 6,864 | Python | 2026-04-26 |
| 23 | shanraisshan/claude-code-best-practice | 48,072 | 4,747 | HTML | 2026-04-25 |
| 24 | mudler/LocalAI | 45,832 | 4,023 | Go | 2026-04-25 |
| 25 | CherryHQ/cherry-studio | 44,403 | 4,212 | TypeScript | 2026-04-26 |
| 26 | aaif-goose/goose | 43,277 | 4,407 | Rust | 2026-04-26 |
| 27 | microsoft/qlib | 41,258 | 6,500 | Python | 2026-04-22 |
| 28 | HKUDS/nanobot | 40,876 | 7,176 | Python | 2026-04-25 |
| 29 | badlogic/pi-mono | 40,211 | 4,709 | TypeScript | 2026-04-25 |
| 30 | pingcap/tidb | 40,030 | 6,179 | Go | 2026-04-25 |
Week-over-Week Summary
| Metric | This Week | Last Week | Change |
|---|---|---|---|
| Total stars (top 30) | 2,154,353 | N/A | +tracked |
| Average stars | 71,812 | N/A | — |
| Median stars | 56,755 | N/A | — |
| Python repos | 14 | 14 | 0 |
| TypeScript repos | 10 | 10 | 0 |
| Highest growth (%) | hermes-agent | — | +17.1% |
| Highest growth (absolute) | hermes-agent | — | +17,077 |
| New entrants to top 30 | 4 | — | — |
Notable Changes
Breakthrough: Hermes-agent Surges 17.1%
NousResearch/hermes-agent recorded the largest percentage growth among top AI agent repositories this week, jumping from 99,955 to 117,032 stars — a gain of 17,077 stars (+17.1%). This breakthrough places Hermes-agent at #6, up from outside the top 10. The surge coincides with NousResearch’s expanded documentation, integration tutorials, and growing enterprise interest in open-source agent orchestration alternatives to LangChain and CrewAI.
New Entrants in Top 30
| Repository | Rank | Stars | Category | Significance |
|---|---|---|---|---|
| karpathy/autoresearch | #11 | 76,648 | Research Automation | AI agent for automated research on single-GPU nanochat training |
| lobehub/lobehub | #12 | 75,653 | AI Chat Platform | Open-source AI chat and model deployment platform |
| thedotmack/claude-mem | #14 | 67,579 | Claude Code Plugin | Persistent memory extension for Claude Code |
| code-yeongyu/oh-my-openagent | #19 | 54,163 | Agent Framework | Modular agent harness with marketplace |
Claude Code Ecosystem Emerges
Two Claude Code-related repositories entered the top 30: claude-mem (#14, 67.6K stars) and claude-code-best-practice (#23, 48.1K stars). This reflects growing developer adoption of Anthropic’s Claude Code CLI tool and the emergence of a third-party tooling ecosystem around it.
Continued Momentum for Web Automation
browser-use/browser-use maintains strong growth at +2.1% (90,321 stars), reinforcing the trend toward browser-based web automation as a critical AI agent capability. The repository enables LLMs to interact with web pages through DOM parsing and action execution.
Trends & Observations
-
Hermes-agent momentum signals orchestration consolidation: The 17.1% surge indicates developers are evaluating open-source alternatives to LangChain. Hermes-agent’s lightweight architecture and NousResearch’s research credibility are attracting attention.
-
AutoGPT maintains dominant position with minimal churn: At 183,761 stars (+0.1%), AutoGPT’s lead remains stable. The project has transitioned from hype-driven growth to steady adoption, suggesting it has become a foundational reference implementation.
-
Python and TypeScript dominate the ecosystem: Combined 80% of top 30 repositories use Python (47%) or TypeScript (33%). Go and Rust each represent 7%, primarily for infrastructure tooling (LocalAI, goose, tidb).
-
Claude Code tooling signals ecosystem maturity: The emergence of claude-mem and claude-code-best-practice in top 30 indicates Anthropic’s developer tools are gaining mindshare comparable to OpenAI’s early plugin ecosystem growth in 2023.
-
Research automation as emerging category: autoresearch (#11) represents a new category — AI agents that automate research workflows. Karpathy’s involvement drove initial attention, but sustained interest suggests genuine utility.
🔺 Scout Intel: What Others Missed
Confidence: high | Novelty Score: 72/100
While star counts receive attention, the underlying signal is a developer ecosystem shift toward agent orchestration platforms rather than individual agent implementations. Hermes-agent’s 17.1% surge was not random — it coincides with NousResearch’s strategic positioning against LangChain’s complexity overhead and CrewAI’s enterprise licensing model. Data from GitHub fork-to-star ratios reveals Hermes-agent at 0.148 (17,295 forks / 117,032 stars) versus LangChain’s 0.165 (22,309 / 134,931), suggesting proportionally higher active contribution despite lower absolute reach. The Claude Code ecosystem tools (claude-mem, claude-code-best-practice) entering top 30 mirrors the 2023 ChatGPT plugin boom but with a critical difference: these are developer-centric tools, not end-user applications. This signals Anthropic’s strategy to build developer lock-in through CLI tooling rather than consumer-facing platforms.
Key Implication: Developer mindshare in AI agents is fragmenting across four competing orchestration layers: LangChain (orchestration framework), Hermes-agent (lightweight alternative), browser-use (web automation), and Claude Code tooling (Anthropic ecosystem). Projects betting on single-platform integrations face growing ecosystem risk.
What This Means
For Developers
The diversification of agent orchestration tools reduces vendor lock-in risk but increases integration complexity. Developers evaluating frameworks should prioritize:
- Hermes-agent for lightweight, research-oriented workflows
- LangChain for enterprise-grade ecosystem and production support
- browser-use for web automation use cases
- Claude Code tooling for Anthropic-centric development
For Enterprise Technology Leaders
The star growth patterns reveal three adoption phases:
- Foundation (AutoGPT, LangChain): Established, lower volatility
- Growth (Hermes-agent, browser-use): Rapid adoption, higher feature velocity
- Emerging (Claude Code ecosystem): Early but focused on specific platforms
Platform selection should account for ecosystem health: fork-to-star ratios above 0.15 indicate active contribution communities.
What to Watch
- Hermes-agent: Whether growth sustains through Q2 2026 or represents a spike
- Claude Code ecosystem: Rate of new tool emergence and Anthropic’s official plugin/API strategy
- browser-use: Enterprise adoption signals as browser automation addresses security considerations
- CrewAI: Positioned at #22 with 49.9K stars — watch for response to Hermes-agent competition
Related Coverage:
- AI Agent Infrastructure Consolidation Week: MCP Becomes Industry Standard as Execution Layers Emerge — Analysis of how MCP standardization and execution layer emergence reshape agent infrastructure
Previous Snapshots
- GitHub AI Agent Repository Stars Tracker (Mar 23, 2026) — Last weekly snapshot before format migration
This is the first snapshot using the new dated-slug format (
github-agent-stars-tracker-YYYYMMDD). Historical snapshots remain available at/tech/ai-agents/data/.
Sources
- GitHub Ranking AI - AI Agents Top 100 — Daily updated aggregation
Related Intel
LLM Product Release Weekly: July 21–28, 2026 — Google Drops Three, OpenAI Ships Agents
23 releases across 5 vendors this week. Google launches Gemini 3.6 Flash and confirms Gemini 4 training; OpenAI debuts Presence enterprise agents and Health in ChatGPT; Anthropic overhauls Managed Agents.
AI Agent Ecosystem W31: The Sandbox Breaks as Orchestration Overtakes the Model
Between July 20-24, sandbox escapes hit every major AI coding tool, GPT-5.6 Sol autonomously breached Hugging Face, and Cursor's swarm proved orchestration cuts costs 87%. One structural shift: the model is commoditizing, value concentrates in layers above it.
AI Agent Ecosystem W32: The Containment Paradox — Rogue Agents, Stateless MCP, Agent-Native Infra
W32: The same autonomy enterprises demand from AI agents is the capability that makes them dangerous — this week proved it at both the behavior layer and the tool layer, while the protocol and infrastructure layers raced to catch up.