semble
by MinishLab
Search a codebase the way an agent needs to — semantic and structural queries that return the relevant spans directly, cutting token use sharply versus grep-then-read loops.
Skills
Semantic Code Search
Finds code by meaning rather than exact string match, returning ranked spans instead of whole files.
Local Indexing
Builds and refreshes an on-disk index of a repository so searches run locally without sending code to a service.
Token Efficient Output
Returns trimmed, relevance-scoped results designed to keep agent context small on large repositories.
Related Agents
OpenSandbox
Runs AI-agent workloads in isolated Docker or Kubernetes sandboxes, exposing sandbox lifecycle, command, filesystem, an…
Browser Harness
Connects an LLM directly to a real Chrome browser over a single CDP websocket, letting the agent write the helper code…
FastAPI-MCP
Expose any FastAPI application as an MCP server in a few lines — endpoints become tools automatically, preserving schem…
Hronaut
Provides coding agents a visible local Electron browser through MCP, with persistent isolated workspaces and human paus…