Connecting Cursor to Dailybot
This article shows you how to connect Cursor to Dailybot so progress reports from your AI workflows show up in Home.
Before you begin
- Install or enable the Dailybot integration Cursor supports for your workspace — typically a Dailybot extension from the marketplace and/or an MCP server that exposes Dailybot tools.
- You need a Dailybot account and an API key or CLI session valid on the machine where Cursor runs.
- Open your project in Cursor so paths to reporter scripts and env files resolve the same way they would in a normal terminal.
Steps
- In Cursor, open Extensions and install the official Dailybot extension if your org uses it, or add the Dailybot MCP server in Cursor Settings under MCP using the connection string or config snippet from the Dailybot docs or dashboard.
- Restart Cursor if the extension or MCP list prompts you to, so the new integration loads.
- Authenticate: run
dailybot loginin Cursor’s integrated terminal, or setDAILYBOT_API_KEYin your shell profile or project.envso agent-run commands inherit it. - Set
DAILYBOT_PROJECT_NAMEin.dailybot.envor.envat the repo root (or use the name field from.devcontainer/devcontainer.jsonif your template relies on that). - Ensure your workspace rules or agent instructions tell the AI to call the reporter after significant work — for example
bash agent_scripts/dailybot-report.sh "message" --metadata '{"model":"your-model-id"}'— matching paths your repo actually contains. - Trigger a test: run the same reporter command once from the integrated terminal with a message like “Cursor Dailybot connection test”.
- In the Dailybot web app, open Home and verify the test entry appears with the expected project name and time.
- If the feed stays empty, confirm the API key, project name, and network access, then use My coding agent’s reports aren’t appearing.
What to expect after
Reports you send from Cursor (manually or via agent instructions) appear in Home alongside other agent activity your org tracks. Cursor-specific UI labels may change between versions; the stable contract is authenticated CLI or API calls from the project environment.