This April 2026 snapshot covers 23 AI agent tools scored by ToolVitals. It gives a grounded view of which projects were shipping well, staying healthy, and looking active enough to matter, without trying to make popularity do all the work.
The ranking is ordered by ToolVitals score first, with GitHub stars used only as a secondary popularity signal when scores are close. That keeps the list centered on product quality and shipping momentum, while still recognizing tools with broader developer attention.
Rankings
| Rank | Tool | Health | Shipping | GitHub Stars | Score | Status |
|---|---|---|---|---|---|---|
| 1 | Agent Deck | 97 | 100 | 2199 | 98 | 🟢 Excellent |
| 2 | Dyad | 95 | 100 | 20186 | 97 | 🟢 Excellent |
| 3 | GoClaw | 92 | 100 | 2938 | 96 | 🟢 Excellent |
| 4 | Moltis | 92 | 100 | 2611 | 96 | 🟢 Excellent |
| 5 | Inkeep Agents | 92 | 100 | 1111 | 96 | 🟢 Excellent |
| 6 | Agent Device | 91 | 100 | 1892 | 95 | 🟢 Excellent |
| 7 | Backend.AI | 89 | 100 | 633 | 94 | 🟢 Excellent |
| 8 | Signet AI | 89 | 100 | 128 | 94 | 🟢 Excellent |
| 9 | Wonda | 86 | 100 | 64 | 92 | 🟢 Excellent |
| 10 | onequery | 83 | 100 | 3 | 91 | 🟢 Excellent |
| 11 | Hermes Agent | 92 | 88 | 121021 | 90 | 🟢 Excellent |
| 12 | Qodo PR Agent | 94 | 72 | 11019 | 84 | 🟢 Excellent |
| 13 | Agent LSP | 81 | 88 | 7 | 84 | 🟢 Excellent |
| 14 | Runtime TypeScript | 81 | 88 | 0 | 84 | 🟢 Excellent |
| 15 | RamaLama | 85 | 72 | 2796 | 79 | 🟢 Good |
| 16 | SkillHub | 85 | 72 | 2665 | 79 | 🟢 Good |
| 17 | Blinko | 82 | 72 | 10140 | 78 | 🟢 Good |
| 18 | GPT Researcher | 81 | 72 | 26739 | 77 | 🟢 Good |
| 19 | Alan AI | 67 | 88 | 2436 | 76 | 🟢 Good |
| 20 | claude-deck | 80 | 72 | 11 | 76 | 🟢 Good |
| 21 | canvas-notebook | 76 | 72 | 8 | 74 | 🟢 Good |
| 22 | Klavis AI | 76 | 38 | 5718 | 59 | 🟡 Fair |
| 23 | Agent Field | 64 | 0 | 726 | 35 | 🔴 Needs Attention |
Top 3 Highlights
Agent Deck ranked first with a 98 ToolVitals score, a 97 health score, and a perfect shipping score. It stands out as a terminal session manager for AI coding agents, with one TUI for Claude, Gemini, OpenCode, Codex, and more.
Dyad came in second with a 97 ToolVitals score and the largest GitHub following in the group at 20,186 stars. It is a local, open-source AI app builder for power users, and it also posted a perfect shipping score.
GoClaw ranked third at 96, with a perfect shipping score and a 92 health score. It is a Go implementation of OpenClaw for multi-tenant agent teams.
Want to see the full health, shipping, openness, and evidence details for every tool in AI Agent Tools? Browse all AI Agent Tools →