$ loading_
Compress, route, remember, and verify AI coding context with major token savings.
The available materials describe an open-source local-binary MCP tool with no required secrets and no declared remote endpoints. Overall risk appears relatively low based on current facts, but it should still be treated as a local privileged tool because it executes code and processes code/context data.
The materials explicitly state that no keys or environment variables are required, and no API tokens, account credentials, or other sensitive secrets are requested; based on current information, credential exposure and abuse risk appears low.
No remote host endpoints are declared, and the materials do not state that user data is sent to external services; based on known facts, no explicit data egress path is identified.
The system flags executes-code, indicating that this MCP tool can execute code or spawn processes locally; this is a standard capability for such tools, but it should still be treated as a local execution surface with constrained runtime permissions.
The description says it 'compresses, remembers, routes, and verifies every token' between code and model, so it can reasonably be inferred that it accesses project context, prompts, or related local data; however, the materials do not specify exact read/write scope, and no clear over-privileged access red flag is evident.
The source is a GitHub open-source repository under Apache-2.0 with about 2.5k stars, providing some community adoption and auditability; unknown maintenance status is a minor uncertainty but not enough on its own to raise this to high risk.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "lean-ctx" yet — see the docs or source repo.
Use lean-ctx to analyze this multi-module repository, compress the context most relevant to the current bug fix, preserve key files, dependencies, and call chains, and output a compact context package ready for a coding model.
A highly relevant, low-token code context summary ready for downstream AI coding.
Use lean-ctx to route context for my task, 'add a retry mechanism to the payment module,' extracting only the needed interfaces, services, tests, and config files, then organize them by priority for Claude Code or Cursor.
A task-ranked context package that reduces irrelevant code and improves assistant focus.
Use lean-ctx to verify whether the code context used by the AI assistant is complete and up to date, check for missing key files or outdated implementations, and provide correction suggestions.
A context completeness and freshness report with actionable correction recommendations.
Read, search, and update codebase documentation context through MCP.
Use a local LLM to index codebases and compress Claude Code context.
Compress Claude Code and OpenCode context losslessly for longer, more efficient coding tasks.
Share code with LLMs via MCP or clipboard with task-specific context rules.
Read, edit, and refactor code precisely with AST-based, token-efficient operations.
Store and retrieve repo patterns so models understand projects instantly.