Per-tool support
Out of 84 analyzed products. Bars scale to the leader; numbers are absolute supportedCount.
Leader: OpenTelemetry with 26 supported products (31.0% of analyzed).
Observability coverage
Which products and frameworks integrate with which observability
stack? Per the May 2026 volumetric agent, observability / debugging is the single highest-demand
question in agent infrastructure (807 HN hits in the
last 12 months, 89% adoption in LangChain's State of Agent
Engineering 2025, six of seven published surveys treat it as a
top-three challenge). This matrix pivots the catalog's eight obs-* columns into a coverage view.
Coverage: 84/912 products (9.2%) have at least one observability cell filled in. Of those analyzed, 52.4% declare support for at least one of the seven tracked tools. The rest of the catalog (mostly research papers and depth-floor rows) is surfaced as unknown by default — the matrix explicitly shows where research has bottomed out rather than papering over the gap.
Legend
- yes documented integration — citation in the cell tooltip
- no no integration / not applicable for this product class
- unknown not yet researched or research bottomed out at the depth-floor; treat as "we don't know"
Matrix — showing 84 of 912
| Product | Tier | Section | LangSmith | OpenTelemetry | Datadog | Helicone | Weave (W&B) | Langfuse | Arize | Coverage |
|---|---|---|---|---|---|---|---|---|---|---|
| Anthropic Claude (foundation models) LangSmith | T1 | Foundation models (substrate reference) | ● | ● | ● | ● | ● | ● | ● | 7 |
| Anthropic Computer Use LangSmith | T1 | Agent frameworks (no first-party memory layer) | ● | ● | ● | ● | ● | ● | ● | 7 |
| Anthropic Computer Use LangSmith | T1 | Agent IDEs & coding harnesses | ● | ● | ● | ● | ● | ● | ● | 7 |
| Gemini Enterprise Agent Platform Memory Bank (rebrand of Vertex AI Memory Bank) LangSmith | T1 | Platform-provider memory | ● | ● | ● | ● | ● | ● | ● | 7 |
| Google Gemini Memory LangSmith | T1 | Platform-provider memory | ● | ● | ● | ● | ● | ● | ● | 7 |
| LangGraph (orchestration) LangSmith | T1 | Agent frameworks (no first-party memory layer) | ● | ● | ● | ● | ● | ● | ● | 7 |
| OpenAI Agents SDK LangSmith | T1 | Agent frameworks (no first-party memory layer) | ● | ● | ● | ● | ● | ● | ● | 7 |
| OpenAI Realtime API LangSmith | T1 | Voice agent platforms | ● | ● | ● | ● | ● | ● | ● | 7 |
| CrewAI LangSmith | T1 | Agent frameworks (no first-party memory layer) | ● | ● | ● | – | ● | ● | ● | 6 |
| Mistral Le Chat — Memories LangSmith | T1 | Platform-provider memory | ● | – | ● | ● | ● | ● | ● | 6 |
| Mistral Vibe (Remote Agents) + Mistral Medium 3.5 LangSmith | T1 | Coding-agent memory | ● | – | ● | ● | ● | ● | ● | 6 |
| Anthropic Claude Memory Datadog | T1 | Platform-provider memory | ○ | – | ● | ● | ● | ● | ● | 5 |
| DeepSeek R1 / V3 family LangSmith | T1 | Foundation models (substrate reference) | ● | – | ● | ● | ● | ● | – | 5 |
| Alibaba Qwen 3 family LangSmith | T1 | Foundation models (substrate reference) | ● | – | ● | – | ● | ● | – | 4 |
| AutoGen / AG2 OpenTelemetry | T1 | Agent frameworks (no first-party memory layer) | – | ● | – | – | ● | ● | ● | 4 |
| Neo4j LangSmith | T1 | Knowledge-graph platforms | ● | ● | ● | ○ | ○ | ○ | ○ | 3 |
| Grok Memory (xAI) LangSmith | T1 | Platform-provider memory | ● | – | – | ● | – | ● | – | 3 |
| OpenHands (All Hands AI) LangSmith | T1 | Agent IDEs & coding harnesses | ● | – | – | ● | – | ● | – | 3 |
| Amazon Neptune Analytics OpenTelemetry | T1 | Knowledge-graph platforms | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| Amazon Q Developer OpenTelemetry | T1 | Agent IDEs & coding harnesses | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| Arize Phoenix OpenTelemetry | T1 | Evaluation & observability platforms | ○ | ● | ○ | ○ | ○ | ○ | ● | 2 |
| Claude Code (Anthropic) OpenTelemetry | T1 | Agent IDEs & coding harnesses | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| Claude Code auto-memory OpenTelemetry | T1 | Claude Code memory mechanisms | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| CLAUDE.md OpenTelemetry | T1 | File-backed / editor paradigms | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| Datadog Bits AI OpenTelemetry | T1 | Use-case-specific agent harnesses | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| Microsoft Copilot Memory OpenTelemetry | T1 | Platform-provider memory | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| v0 (Vercel) OpenTelemetry | T1 | Agent IDEs & coding harnesses | ○ | ● | ● | ○ | ○ | ○ | ○ | 2 |
| LanceDB LangSmith | T1 | Vector-database infrastructure | ● | ● | – | ○ | ○ | ○ | ○ | 2 |
| OpenAI Codex CLI OpenTelemetry | T1 | Agent IDEs & coding harnesses | ○ | ● | – | ● | – | – | – | 2 |
| Mem0 LangSmith | T1 | Dedicated memory layers | ● | – | – | – | – | ● | – | 2 |
| Braintrust OpenTelemetry | T1 | Evaluation & observability platforms | ○ | ● | ○ | ○ | ○ | ○ | ○ | 1 |
| Perplexity Deep Research Helicone | T1 | Use-case-specific agent harnesses | ○ | ○ | ○ | ● | ○ | ○ | ○ | 1 |
| Hugging Face LeRobot Weave (W&B) | T1 | Robotics foundation models & agent stacks | ○ | – | ○ | ○ | ● | ○ | ○ | 1 |
| Salesforce Agentforce Datadog | T1 | Use-case-specific agent harnesses | ○ | – | ● | ○ | ○ | ○ | ○ | 1 |
| NVIDIA GR00T / Isaac Weave (W&B) | T1 | Robotics foundation models & agent stacks | ○ | – | – | ○ | ● | ○ | ○ | 1 |
| Browser Use LangSmith | T1 | Computer-use & desktop agents | ● | – | – | – | – | – | – | 1 |
| Cline (framework) Helicone | T1 | Agent frameworks (no first-party memory layer) | – | – | – | ● | – | – | – | 1 |
| Cognee Langfuse | T1 | Dedicated memory layers | – | – | – | – | – | ● | – | 1 |
| Continue.dev (framework) OpenTelemetry | T1 | Agent frameworks (no first-party memory layer) | – | ● | – | – | – | – | – | 1 |
| ElevenLabs Conversational AI Langfuse | T1 | Voice agent platforms | – | – | – | – | – | ● | – | 1 |
| Graphiti MCP Server (Zep) LangSmith | T1 | Claude Code memory mechanisms | ● | – | – | – | – | – | – | 1 |
| Letta / MemGPT OpenTelemetry | T1 | Dedicated memory layers | – | ● | – | – | – | – | – | 1 |
| Sourcegraph Cody (agent) OpenTelemetry | T1 | Agent IDEs & coding harnesses | – | ● | – | – | – | – | – | 1 |
| Zep & Graphiti LangSmith | T1 | Dedicated memory layers | ● | – | – | – | – | – | – | 1 |
| Abridge | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| AGENTS.md | T1 | File-backed / editor paradigms | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Bolt.new (StackBlitz) — harness | T1 | Agent IDEs & coding harnesses | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Character.ai | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| ChatGPT — Codex (cloud agent) | T1 | Agent IDEs & coding harnesses | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Claude for Chrome | T1 | Computer-use & desktop agents | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| claude-mem | T1 | Claude Code memory mechanisms | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Cognition Devin v2 / Spec Mode | T1 | Agent IDEs & coding harnesses | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Decagon | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Devin (Cognition) | T1 | Coding-agent memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Engram (DeepSeek) | T1 | Recent method papers — theorized, no distinct product | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Google DeepMind Gemini Robotics | T1 | Robotics foundation models & agent stacks | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Granola | T1 | Voice-first / wearable AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Harvey Memory | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Hippocratic AI Polaris | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| HubSpot Breeze AI Agents | T1 | Use-case-specific agent harnesses | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| LinkedIn Cognitive Memory Agent | T1 | Platform-provider memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Lovable.dev — harness | T1 | Agent IDEs & coding harnesses | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Meta AI Memory | T1 | Platform-provider memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Microsoft Edge Copilot Mode + Journeys | T1 | Browser-agent memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| NVIDIA Nemotron 3 | T1 | Recent method papers — theorized, no distinct product | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| OpenAI ChatGPT Memory | T1 | Platform-provider memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| OpenAI Deep Research | T1 | Use-case-specific agent harnesses | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| OpenAI Operator | T1 | Computer-use & desktop agents | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Perplexity Comet | T1 | Browser-agent memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Physical Intelligence (π) | T1 | Robotics foundation models & agent stacks | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Project Mariner | T1 | Computer-use & desktop agents | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Sierra | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Suki AI | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| Wayve GAIA-2 / GAIA-3 | T1 | Vertical / domain-specific AI memory | ○ | ○ | ○ | ○ | ○ | ○ | ○ | 0 |
| GraphRAG (Microsoft) | T1 | Retrieval-as-memory hybrids | ○ | – | ○ | ○ | ○ | ○ | ○ | 0 |
| MCP Servers (Reference) | T1 | Agent frameworks (no first-party memory layer) | ○ | – | ○ | ○ | ○ | ○ | ○ | 0 |
| Model Context Protocol (MCP spec) | T1 | Agent frameworks (no first-party memory layer) | ○ | – | ○ | ○ | ○ | ○ | ○ | 0 |
| Official MCP Memory server | T1 | Claude Code memory mechanisms | ○ | – | ○ | ○ | ○ | ○ | ○ | 0 |
| Anthropic Auto Dream | T1 | Claude Code memory mechanisms | ○ | – | – | ○ | ○ | ○ | ○ | 0 |
| GitHub Copilot Workspace | T1 | Agent IDEs & coding harnesses | ○ | – | – | ○ | ○ | ○ | ○ | 0 |
| JetBrains AI Assistant | T1 | Agent IDEs & coding harnesses | ○ | – | – | ○ | ○ | ○ | ○ | 0 |
| Lucidworks Conversational Q&A AI Agent | T1 | Enterprise-search adjacencies | ○ | – | – | ○ | ○ | ○ | ○ | 0 |
| Sinequa | T1 | Enterprise-search adjacencies | ○ | – | – | ○ | ○ | ○ | ○ | 0 |
| Cursor Rules | T1 | File-backed / editor paradigms | ○ | – | – | – | – | – | – | 0 |