Sourcegraph (Amp)

1. 

   Code Search

   Executes literal, keyword, and regular expression searches across an entire codebase in milliseconds. Filters results by file paths, programming languages, commits, or custom patterns to deliver exact matches across multiple repositories.

   Use caseAn application section leader locates where a specific code snippet is used across multiple repositories in seconds without asking other engineers.

   From sourcegraph.com/code-search · image from Sourcegraph
2. 

   Code Navigation

   Jumps to symbol definitions and references across multiple repositories directly in a web browser. Relies on semantic index analysis and falls back to syntactic or search based methods when needed.

   Use caseA developer traces cross repository symbol definitions and references directly in the browser without cloning every repository locally.

   From sourcegraph.com/code-search · image from Sourcegraph
3. MCP Server

   Supplies whole codebase context and semantic index data to external AI coding agents. Enables agents to plan and execute multi repository edits with lower token spend and fewer retries.

   Use caseAn engineering team connects internal AI coding agents to an enterprise codebase index to prevent missing authentication guards or frontend routes during cross cutting refactors.

   From sourcegraph.com
4. Agentic Batch Changes

   Applies large scale code updates across all repositories using an AI agent. Scopes the planned update, implements code modifications, and monitors changes through to the final merge.

   Use caseAn engineering team migrates deprecated syntax or remediates a security risk across every repository simultaneously.

   From sourcegraph.com

Written from Sourcegraph's own website, checked Oct 16, 2026. Each feature links to the page it comes from; images are Sourcegraph's own.