AI Agent (MCP)
Let Cursor, VS Code Copilot, or any MCP client read and resolve reports directly.
The AI Agent connects your AI coding assistant to SnagRelay over the Model Context Protocol (MCP), so it can pull full bug report context — screenshot, environment, console/network data, reproduction steps — straight into a chat, and mark reports resolved once fixed.
Setting it up
Generate a token
Go to AI Agent Setup on your Project and generate an MCP token.
Connect your editor
The setup screen provides ready-to-paste config snippets for Cursor and VS Code
Copilot. If you use the VS Code / Cursor extension,
signing in there writes this config automatically (.vscode/mcp.json or
.cursor/mcp.json).
Use it from chat
Once connected, ask your assistant things like "show me open bug reports" or "start fixing the checkout bug" — it calls the underlying MCP tools directly.
Available MCP tools
| Tool | Purpose |
|---|---|
list_reports | List open reports for a project |
get_report | Fetch full detail for one report |
start_report | Mark a report in progress (optionally creating a branch) |
set_report_branch | Associate a git branch with a report |
mark_report_resolved | Mark a report resolved |
These power the VS Code / Cursor extension — starting or resolving a report from the extension calls these tools directly, there's no manual drag-and-drop step involved.
Revoking access
Revoke an MCP token at any time from AI Agent Setup — this immediately invalidates any editor connections using it.