LeanCTX documentation

LeanCTX is the Context Gateway for AI systems.

It sits between AI and the systems it reads. It selects relevant context, applies supported controls, and records what the integration can observe.

Select. Control. Prove.

Select sources. Shape the view. Check supported reads before delivery.

Connect your AI tool.

Install on your system, connect your AI tool and verify your first read.

Embed supported context selection, controls and evidence. Your host owns model calls; Python continuity handoffs remain Preview.

See the difference

Start with the API surface. Recover the full implementation when the task needs it.

With LeanCTX installed, replace the path with a file in your project.

Read. Inspect. Recover.CLI

# Inspect the API surface
lean-ctx read src/main.rs -m signatures

# Read the full source
lean-ctx read src/main.rs -m full --fresh

Two views of the same file. On Windows, macOS or Linux.

Learn the essentials

Connect evidence, control input and execution, then evaluate and reuse the result.

Find the right source. Choose the detail your task needs.

Set access boundaries and configure read-time checks.

Carry selected knowledge into the next task.

Inspect delivery, token reduction and their limits.

Set allowed routes, budgets, retries and organizational policies.

Local task ownership and bounded work graphs support coordination. Broader multi-agent orchestration remains a Research capability.

Inspect local activity, context and project signals.

Prepare your codebase for lexical, graph and semantic retrieval.

Go straight to the source

Find exact schemas and versioned runtime help alongside the shared Engine concepts.