Semantic code-search and context layer that turns repos into token-optimized markdown for AI agents and human engineers.
What it does
Code Fundi is a codebase context layer that indexes public and private repositories and distills them into logic-dense, token-optimized markdown. It offers sub-second semantic and AST-aware search across many codebases, cutting the boilerplate noise that inflates AI token usage, and plugs into agents and IDEs via MCP and API.
Core features
Semantic and AST-aware code search
Repo-to-markdown distillation
Multi-repo search
MCP/API plus IDE extensions (VS Code, Cursor, Windsurf)
Training-data generation for fine-tuning
Real-time codebase observability
Zero-data-retention option
Best for
→Ground AI coding agents with codebase context
→Reduce agent token consumption
→Search patterns across many repos
→Generate fine-tuning datasets
→Legacy modernization and due diligence
Pricing
Free
Dev
$21/mo
Pro
$210/mo
Toolspool rankingby monthly traffic