DeepEval for AI coding agents
Testing & code quality · deepeval.com
Official Anthropic-marketplace plugin with 3 skills (no bundled MCP server).
Skills for adding DeepEval evaluations, tracing, datasets, Confident AI reports, and iterative improvement loops to AI applications.
How DeepEval works in each agent
Scores appear only after we verify them, with a source and date. Methodology
| Agent | Score | MCP | Remote / OAuth | Skill / plugin | Verified |
|---|---|---|---|---|---|
| C Claude Code | 4/5 | No | Unknown | Yes | Oct 16, 2026 |
| X OpenAI Codex | Not yet verified. Testing in progress. | ||||
| ▲ Cursor | Not yet verified. Testing in progress. | ||||
| G GitHub Copilot | Not yet verified. Testing in progress. | ||||
| A Google Antigravity | Not yet verified. Testing in progress. | ||||
Setup by agent
/plugin install deepeval@claude-plugins-officialWe haven't verified a OpenAI Codex setup for DeepEval yet. If the vendor documents an MCP server, the MCP config generator gives you the right ~/.codex/config.toml (user) format. Config locations: Codex setup.
We haven't verified a Cursor setup for DeepEval yet. If the vendor documents an MCP server, the MCP config generator gives you the right .cursor/mcp.json (project) format. Config locations: Cursor setup.
We haven't verified a GitHub Copilot setup for DeepEval yet. If the vendor documents an MCP server, the MCP config generator gives you the right .vscode/mcp.json (workspace) format. Config locations: Copilot setup.
We haven't verified a Google Antigravity setup for DeepEval yet. If the vendor documents an MCP server, the MCP config generator gives you the right ~/.gemini/config/mcp_config.json (global) format. Config locations: Antigravity setup.
Evidence
- ↗https://claude.com/marketplace/plugins
- ↗https://github.com/anthropics/claude-plugins-official/blob/main/.claude-plugin/marketplace.json
- ↗https://github.com/confident-ai/deepeval
- ↗https://github.com/confident-ai/deepeval
Bundled skills (3)
deepevaldeepeval-oteldeepeval-tracing Checked plugin files at https://github.com/confident-ai/deepeval (.mcp.json / plugin.json mcpServers; skills/*/SKILL.md), Oct 16 2026
Agent Stack Weekly · free
What changed across coding agents this week
Limit and pricing changes, new MCP servers, and setup fixes for Claude Code, Codex, Cursor, Copilot and Antigravity.
No spam. Unsubscribe anytime. Read past issues