Sourcegraph for AI agents
Sourcegraph is a code search and intelligence platform that lets developers search, browse, and compare code across many repositories at once. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Sourcegraph.
Search across every repository
Search repository files for a function or pattern across every codebase indexed, not just one repo.
Compare two commits before merging
Compare two commits to see exactly what changed before an agent recommends approving a pull request.
Pull file contents for review
Get the contents of a specific file at a commit, read-only, to answer a question about the code.
8 Sourcegraph actions, graded by risk.
Every Sourcegraph action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
8Look things up. Allowed by default.
- sourcegraph.get_current_userGet Current User
- sourcegraph.get_file_contentsGet File Contents
- sourcegraph.list_repositoriesList Repositories
- sourcegraph.get_commit_detailsGet Commit Details
- sourcegraph.list_repository_filesList Repository Files
- sourcegraph.list_repository_languagesList repository languages
- sourcegraph.compare_commitsCompare Commits
- sourcegraph.check_site_settings_edit_permissionCheck Site Settings Edit Permission
write
0Create and change things. Allow, or ask the user first.
- No write actions.
destructive
0Delete, cancel or archive. Ask first, or deny outright.
- No destructive actions.
Sourcegraph in three steps.
- 01Your users connect SourcegraphThey add their Sourcegraph api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, and writes like “updates” can wait for the user to approve.
- 03Any agent can actYour agent calls Sourcegraph through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('sourcegraph', { read: 'allow', write: 'ask', destructive: 'deny', }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Sourcegraph connects.
Users add their Sourcegraph api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Sourcegraph connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Sourcegraph from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Sourcegraph and Arc0, answered.
Can I use Sourcegraph with Claude, ChatGPT or Cursor?
Yes. Connect Sourcegraph to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Sourcegraph actions you allow.
How do users connect Sourcegraph?
Users add their Sourcegraph api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Sourcegraph actions can my agent take?
8 in total: 8 read, 0 write and 0 destructive. Your policies decide which of them each agent may call.
Can I make my agent read-only in Sourcegraph?
Yes. Allow read actions and deny writes in the Sourcegraph policy. Your agent can still look things up, and any write it attempts is blocked and logged.
Can my own backend call Sourcegraph too?
Yes. The same Sourcegraph connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Sourcegraph into your agent.
Your users connect Sourcegraph once, under your brand. Your agent gets 8 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan