AI Agent Trust Snapshot — Week 22, 2026
Every week HVTracker diffs its registry snapshots and publishes what actually changed: trust-score moves of 3+ points, new listings, and gained provenance or MCP support. This edition compares the 2026-05-24 and 2026-05-31 snapshots — 79 gains, 0 declines, 113 new listings, 1 provenance and 0 MCP additions. Every figure below is generated directly from those snapshots.
Trust score up (79)
| Agent | HVTrust | Change |
|---|---|---|
| LangGraph | 0 → 93.3 | +93.3 |
| Codex | 0 → 92.8 | +92.8 |
| Vercel AI SDK | 0 → 91.9 | +91.9 |
| PydanticAI | 0 → 91.9 | +91.9 |
| Haystack | 0 → 90.2 | +90.2 |
| LiveKit Agents | 0 → 89.9 | +89.9 |
| Trigger.dev | 0 → 89.8 | +89.8 |
| Cline | 0 → 89.0 | +89.0 |
| Strands Agents | 0 → 88.5 | +88.5 |
| Mem0 | 0 → 87.6 | +87.6 |
| CrewAI | 0 → 81.0 | +81.0 |
| Graphiti | 0 → 79.8 | +79.8 |
| AutoGen | 0 → 79.1 | +79.1 |
| LiteLLM | 0 → 77.3 | +77.3 |
| OpenHands | 0 → 75.2 | +75.2 |
| Gemini CLI | 0 → 74.9 | +74.9 |
| Dify | 0 → 73.4 | +73.4 |
| Firecrawl | 0 → 73.0 | +73.0 |
| Agno | 0 → 72.5 | +72.5 |
| CopilotKit | 0 → 72.3 | +72.3 |
| Flowise | 0 → 72.2 | +72.2 |
| E2B | 0 → 72.1 | +72.1 |
| Composio | 0 → 71.7 | +71.7 |
| Browser Use | 0 → 71.2 | +71.2 |
| CAMEL | 0 → 71.1 | +71.1 |
| DeerFlow | 0 → 70.7 | +70.7 |
| AutoGPT | 0 → 70.6 | +70.6 |
| GPT Researcher | 0 → 70.3 | +70.3 |
| Semantic Kernel | 0 → 70.2 | +70.2 |
| AgentScope | 0 → 69.8 | +69.8 |
| Google ADK (Python) | 0 → 69.4 | +69.4 |
| Flyte | 0 → 68.9 | +68.9 |
| GPTMe | 0 → 68.4 | +68.4 |
| Qwen Code | 0 → 68.1 | +68.1 |
| Cognee | 0 → 68.1 | +68.1 |
| Open SWE | 0 → 67.2 | +67.2 |
| Microsoft Agent Framework | 0 → 66.5 | +66.5 |
| Khoj | 0 → 66.1 | +66.1 |
| Google ADK (Go) | 0 → 65.4 | +65.4 |
| WrenAI | 0 → 65.1 | +65.1 |
| Cua | 0 → 64.9 | +64.9 |
| Open Interpreter | 0 → 64.7 | +64.7 |
| Guidance | 0 → 64.2 | +64.2 |
| VoltAgent | 0 → 63.7 | +63.7 |
| SWE-agent | 0 → 61.7 | +61.7 |
| Aider | 0 → 60.8 | +60.8 |
| Maxun | 0 → 60.7 | +60.7 |
| Spring AI Alibaba | 0 → 60.4 | +60.4 |
| UFO | 0 → 59.3 | +59.3 |
| Roo Code | 0 → 58.8 | +58.8 |
| ChatDev | 0 → 58.6 | +58.6 |
| AdalFlow | 0 → 58.5 | +58.5 |
| GitHub Copilot CLI | 0 → 58.1 | +58.1 |
| UI-TARS Desktop | 0 → 57.8 | +57.8 |
| Letta | 0 → 56.0 | +56.0 |
| Coze Studio | 0 → 55.1 | +55.1 |
| Browser Use Web UI | 0 → 53.8 | +53.8 |
| Qwen Agent | 0 → 50.5 | +50.5 |
| Tabby | 0 → 50.2 | +50.2 |
| OmniParser | 0 → 48.1 | +48.1 |
| AnythingLLM | 0 → 47.3 | +47.3 |
| GPT Pilot | 0 → 46.6 | +46.6 |
| MetaGPT | 0 → 46.2 | +46.2 |
| UI-TARS | 0 → 44.3 | +44.3 |
| RA.Aid | 0 → 43.7 | +43.7 |
| Leon | 0 → 42.0 | +42.0 |
| Activepieces | 0 → 41.7 | +41.7 |
| Sweep | 0 → 40.8 | +40.8 |
| PromptFlow | 0 → 39.8 | +39.8 |
| Devika | 0 → 38.8 | +38.8 |
| PraisonAI | 0 → 37.7 | +37.7 |
| Eliza | 0 → 37.0 | +37.0 |
| AgentOps | 0 → 35.2 | +35.2 |
| Continue | 0 → 32.5 | +32.5 |
| BabyAGI | 0 → 31.9 | +31.9 |
| TEN Framework | 0 → 25.0 | +25.0 |
| Plandex | 0 → 23.6 | +23.6 |
| Devon | 0 → 22.4 | +22.4 |
| TaskWeaver | 0 → 19.9 | +19.9 |
Newly listed (113)
| Agent | Category | HVTrust |
|---|---|---|
| OpenAI Agents SDK | Agent Frameworks | 91.7 |
| n8n | Workflow Platforms | 91.0 |
| LeRobot | Robotics & Embodied | 88.8 |
| Skyvern | Browser & Computer Use | 88.4 |
| A2A / Agent2Agent Protocol | Protocols & Tool Integration | 88.3 |
| LanceDB | Memory & Knowledge | 88.0 |
| Vespa | Memory & Knowledge | 87.2 |
| Stagehand | Browser & Computer Use | 87.1 |
| Open WebUI | Memory & Knowledge | 86.3 |
| Botpress | Workflow Platforms | 85.8 |
| Pipecat | Voice & Conversational | 85.6 |
| Weights & Biases Weave | Observability & Evaluation | 84.1 |
| LangChain | Agent Frameworks | 78.6 |
| Langfuse | Observability & Evaluation | 78.1 |
| AIPass | Agent Frameworks | 77.0 |
| Promptfoo | Observability & Evaluation | 76.6 |
| Weaviate | Memory & Knowledge | 76.2 |
| Arize Phoenix | Observability & Evaluation | 75.9 |
| Qdrant | Memory & Knowledge | 74.6 |
| NeMo Guardrails | Security & Guardrails | 73.3 |
| Goose | Coding Agents | 73.0 |
| Milvus | Memory & Knowledge | 72.8 |
| Chroma | Memory & Knowledge | 72.7 |
| OpenClaw | Browser & Computer Use | 72.3 |
| Langflow | Workflow Platforms | 71.8 |
| Giskard | Observability & Evaluation | 70.7 |
| LangWatch | Observability & Evaluation | 69.8 |
| Garak | Security & Guardrails | 68.4 |
| RAGFlow | Memory & Knowledge | 68.1 |
| LibreChat | Memory & Knowledge | 66.8 |
| Griptape | Agent Frameworks | 66.8 |
| DeepEval | Observability & Evaluation | 66.5 |
| MCP Registry | Protocols & Tool Integration | 65.5 |
| Burr | Agent Frameworks | 65.0 |
| Model Context Protocol / MCP | Protocols & Tool Integration | 64.7 |
| Marvin | Agent Frameworks | 63.6 |
| Guardrails AI | Security & Guardrails | 62.7 |
| Claude Code | Coding Agents | 61.9 |
| txtai | Memory & Knowledge | 61.8 |
| Crawl4AI | Research & Data | 61.3 |
| Agency Swarm | Agent Frameworks | 61.1 |
| Rasa | Workflow Platforms | 61.1 |
| Langroid | Agent Frameworks | 60.9 |
| Atomic Agents | Agent Frameworks | 59.3 |
| Helicone | Observability & Evaluation | 58.9 |
| BrowserGym agents | Browser & Computer Use | 58.5 |
| Ragas | Observability & Evaluation | 56.4 |
| Llama Guard | Security & Guardrails | 55.2 |
| TinyTroupe | Agent Frameworks | 54.5 |
| Swarm by OpenAI | Agent Frameworks | 52.8 |
| pgvector | Memory & Knowledge | 52.7 |
| FastAgency | Agent Frameworks | 52.1 |
| LLM Guard | Security & Guardrails | 50.5 |
| CrewAI tools | Protocols & Tool Integration | 49.3 |
| Mastra | Agent Frameworks | 49.0 |
| Superagent | Workflow Platforms | 47.8 |
| LlamaIndex | Agent Frameworks | 47.4 |
| ControlFlow | Agent Frameworks | 47.3 |
| opencode | Coding Agents | 46.6 |
| PrivateGPT | Memory & Knowledge | 46.5 |
| Agent-E | Browser & Computer Use | 46.3 |
| Daytona | Sandboxes & Runtimes | 45.9 |
| Google Genkit | Agent Frameworks | 44.8 |
| Sotopia agents | Agent Frameworks | 44.0 |
| Quivr | Memory & Knowledge | 43.8 |
| OpenCUA | Browser & Computer Use | 43.1 |
| STORM | Research & Data | 42.3 |
| smolagents | Agent Frameworks | 41.9 |
| TruLens | Observability & Evaluation | 41.9 |
| Agenta | Observability & Evaluation | 41.2 |
| Bee Agent Framework | Agent Frameworks | 38.3 |
| JARVIS / HuggingGPT | Agent Frameworks | 37.4 |
| OpenSSA | Agent Frameworks | 37.0 |
| Open Operator | Browser & Computer Use | 36.3 |
| SwarmZero | Agent Frameworks | 35.4 |
| AgentLite | Agent Frameworks | 34.3 |
| Agent Laboratory | Agent Frameworks | 34.0 |
| Rebuff | Security & Guardrails | 29.7 |
| OpenAgents | Agent Frameworks | 29.4 |
| Voyager Minecraft Agent | Agent Frameworks | 28.3 |
| Mentat | Coding Agents | 27.7 |
| CodeAct | Coding Agents | 27.2 |
| PearAI | Coding Agents | 25.9 |
| XAgent | Agent Frameworks | 25.7 |
| AgentLLM | Agent Frameworks | 25.4 |
| TinyAgents | Agent Frameworks | 25.1 |
| AIlice | Agent Frameworks | 25.0 |
| OpenManus | Agent Frameworks | 24.4 |
| Open Deep Research | Research & Data | 24.2 |
| Generative Agents | Agent Frameworks | 24.0 |
| WebVoyager | Browser & Computer Use | 22.7 |
| Agent Hospital | Agent Frameworks | 19.7 |
| OpenAGI | Agent Frameworks | 13.5 |
| Concordia | Multi-Agent Systems | 0.0 |
| Nx | Agent Frameworks | 0.0 |
| Traceroot | LLM Gateways & Infra | 0.0 |
| Agent Governance Toolkit | Agent Frameworks | 0.0 |
| Tidb | Memory & Knowledge | 0.0 |
| Infracost | Agent Frameworks | 0.0 |
| 1Panel | Agent Frameworks | 0.0 |
| Verl Agent | Agent Frameworks | 0.0 |
| Goose | Agent Frameworks | 0.0 |
| Neo | Memory & Knowledge | 0.0 |
| Bernstein | Coding Agents | 0.0 |
| Budibase | Workflow Platforms | 0.0 |
| Materialize | Agent Frameworks | 0.0 |
| Pyod | Agent Frameworks | 0.0 |
| Holmesgpt | LLM Gateways & Infra | 0.0 |
| Airbyte | Workflow Platforms | 0.0 |
| Lumibot | Multi-Agent Systems | 0.0 |
| Webarena | Agent Frameworks | 0.0 |
| Fedml | Agent Frameworks | 0.0 |
| Nocobase | Workflow Platforms | 0.0 |
Provenance gained (1)
| Agent | Category |
|---|---|
| Codex | Coding Agents |
Generated automatically from the 2026-05-24 and 2026-05-31
registry snapshots — no numbers in this post are estimated or written by hand.
Method: HVTrust methodology ·
live diff: what changed this week ·
feed: RSS.