Integrations
Use DevTune with your agent
Connect your agent to your DevTune project and it can read what AI platforms say about your product, pick the next fix and check whether a shipped fix moved the numbers. Choose your agent for its connect steps.
13 agents with a setup guide
Coding agents
- Claude CodeMade byAnthropicConnect withSign-in, API key, SkillsHow we checkedRun with DevTune
Coding agent for the terminal, desktop app and IDEs.
- CodexMade byOpenAIConnect withSign-in, API key, SkillsHow we checkedRun with DevTune
Coding agent in the terminal, IDE extension and ChatGPT desktop app.
- Muse CodeMade byMetaConnect withSign-in, API key, SkillsHow we checkedRun with DevTune
Coding agent for the terminal.
- CursorMade byAnysphereConnect withSign-in, API keyHow we checkedConnected to DevTune
AI code editor with a command-line agent.
- Gemini CLIMade byGoogleConnect withSign-in, API keyHow we checkedChecked against docs
Coding agent for the terminal.
- OpenCodeMade bySSTConnect withSign-in, API keyHow we checkedChecked against docs
Open-source coding agent for the terminal.
- PiMade byEarendil WorksConnect withSign-in, API key, SkillsHow we checkedChecked against docs
Minimal coding agent for the terminal.
- DroidMade byFactoryConnect withSign-in, API keyHow we checkedChecked against docs
Coding agent for the terminal.
- Devin CLIMade byCognitionConnect withSign-in, API keyHow we checkedChecked against docs
Coding agent for the terminal, with cloud sessions.
Chat products
Personal agents
- Hermes AgentMade byNous ResearchConnect withSign-in, API keyHow we checkedChecked against docs
Open-source personal agent you run yourself.
Agent frameworks
- EveMade byVercelConnect withAPI keyHow we checkedChecked against docs
Framework for building durable backend agents.
Run with DevTune means the agent connected to DevTune and called its tools. Connected to DevTune means it connected and listed the tools. Checked against docs means every step comes from the maker's current documentation or the agent's own help, and was not run.
What your agent connects to
| Route | Use it when | Reference |
|---|---|---|
MCP server at https://devtune.ai/api/mcp | Your agent supports remote MCP servers. It signs in through the browser or sends a project API key. | MCP server |
| DevTune skills | Your agent loads agent skills. They teach it which DevTune tool answers which question and how to report the numbers. | Agent skills |
devtune CLI | Your agent can run shell commands but has no MCP support. | CLI |
REST API at https://devtune.ai/api/v2 | Your agent can make HTTP requests, or you are building the workflow yourself. | API overview |
Every route reaches one project.
Any other agent
Your agent is not listed? Use the first route it supports:
- Remote MCP server. Add a Streamable HTTP server at
https://devtune.ai/api/mcp. If the agent can sign in to MCP servers, sign in and select your project. If not, send a project API key in theAuthorization: Bearerheader. MCP server has the generic configuration. - Shell access. Install the CLI, set
DEVTUNE_API_KEYandDEVTUNE_PROJECT_ID, and have the agent rundevtune instructionsto read how to use it. - HTTP requests. Give the agent a project API key and the OpenAPI description.
No DevTune account yet
Your agent can register, draft the project and hand you a code to confirm. See Set up DevTune with an AI agent.
Make AI describe you the way you'd describe yourself.
No card required