Who it is for
Teams building AI agents who already keep their test sets and metrics in Coval.
Overview
What Coval does
Coval is an evaluation platform for AI agents, and this app exposes the objects inside it to a conversation. Test sets, personas, metrics, agents, and past evaluation runs can all be inspected.
Beyond reading, evaluation resources can be created and refined, and evaluation runs can be launched from the same place. That covers the setup work as well as the execution.
Analysis is handled separately. An assistant named Sofia answers questions read-only, reporting on what sits in the connected Coval organization without altering it. The record names no example prompts or bundled skills.
What stands out
Analysis runs through a read-only assistant that is kept apart from the tools which create or launch evaluations.
Availability
Where Coval is listed
Coval ships this on both marketplaces: Coval is listed in the Claude connectors directory and in the ChatGPT apps directory, under the same publisher domain.
Only 392 of the 3,150 products in this directory are listed in both marketplaces. One MCP server can serve both surfaces; what differs is each platform's listing and review process.
ChatGPT listing
In the ChatGPT apps directory
- Developer
- Coval
- Filed under
- Developer Tools
- Listed since
- July 2026
- Authentication
- Authenticates when installed
- Website
- coval.ai
- Policies published
- Privacy policy · Terms of service
- Official listing
- Open in ChatGPT
- Snapshot
- August 2026
Claude listing
In the Claude connectors directory
Coval publishes 19 tools to Claude, deeper than 67.2% of the 1,253 Claude-listed connectors.
The median Claude-listed connector filed under Developer Tools publishes 13, so this one runs 6 tools above its bucket's median and ranks 37th of 95 by tool count.
It falls in the 11-20 tools band, which holds 22.5% of the Claude directory.
Coval publishes a remote MCP endpoint. Claude connects to it over HTTP; there is nothing to install locally.
Every connector in the Claude directory is remote, all 1,253 of them, so transport is not a differentiator there.
- Publisher
- Coval
- Verification
- Community-listed (61.5% of the Claude directory)
- Filed under
- Developer Tools
- Endpoint
-
https://mcp.coval.dev/claude/mcp - Documentation
- Publisher setup docs
- Published tools
- 19 · 11-20 tools band
- Depth rank in Developer Tools
- 37 of 95
- Snapshot
- August 2026
Coval is community-listed in the Claude directory, the tier 770 of 1,253 connectors sit in (61.5%). It is the default for a self-published server, and says nothing either way about quality.
In the publisher's words
How Coval describes it
Connect Claude to your Coval workspace to inspect agents, test sets, test cases, personas, metrics, and evaluation runs. Build or update evaluation fixtures, launch safe chat and SMS evaluation runs, and ask Sofia for grounded, read-only analysis and recommendations. The Claude connector does not start voice, outbound voice, or WebSocket voice runs.
Quoted from the connector's own Claude directory listing as captured in August 2026. Node8 did not write it and does not vouch for it.
Nearby
Other Developer Tools connectors
Botpress
Agent prototyping on the Botpress platform
ChatGPT
Build Web Data Visualization
Design and build browser-based charts and dashboards
ChatGPT
Buildy
Hosted runtime for small apps agents create
ChatGPT
CircleCI
CI configuration and build troubleshooting inside Codex
ChatGPT
ElevenLabs
Create and configure ElevenLabs voice agents
Claude
Mem0
Persistent memory that survives between Claude sessions
Claude
This is an independent catalogue entry compiled by Node8 from Coval's public marketplace listings as they stood in August 2026. Listing facts come from the marketplaces; the comparisons against the rest of the directory and any live-endpoint measurements are Node8's. Node8 is not affiliated with Coval, and inclusion here is not an endorsement.
Want your company reachable inside ChatGPT and Claude?
Node8 builds MCP servers end to end: deciding which tools are worth exposing, the authentication and permission model, the review submissions, and the onboarding that gets them used. One server serves ChatGPT, Claude, and Microsoft Copilot.