OpenWiki is an open-source agent harness developed by LangChain designed to generate, ground, and continuously maintain codebase documentation and internal wikis. It orchestrates agentic workflows that analyze repository structures, code diffs, and commit histories to keep developer documentation synchronized with evolving software projects. In daily workflows, OpenWiki operates via a command-line interface (such as running update commands against branches) to inspect repository states and generate or refresh markdown documentation. The system supports multi-model architectures, enabling developers to run maintenance workflows using leading LLM providers, including Anthropic Claude and Google Gemini models. What distinguishes OpenWiki from general-purpose coding agents is its specialized focus on longitudinal documentation grounding, factual accuracy, and knowledge retention. It includes built-in benchmarking systems (such as the LEDGER longitudinal evaluation framework) to evaluate wiki grounding, monitor factual precision, and mitigate knowledge forgetting as codebases evolve over long commit histories. OpenWiki is distributed as open-source software under LangChain's GitHub organization, making it readily embeddable into local development workflows and continuous integration pipelines.
Continuous Wiki Updates — Executes automated synchronization workflows to update repository documentation in response to code changes. - LEDGER Longitudinal Benchmark — Built-in evaluation harness for testing documentation grounding, maintenance rates, and fact retention across commit histories. - Multi-Model Architecture — Supports integration with diverse LLMs, including Claude and Gemini models, for bounded extraction and generation tasks. - CLI-Driven Orchestration — Provides direct terminal entrypoints to trigger documentation updates and evaluation passes. - Bounded Factual Precision Judges — Employs structured model evaluator pipelines to verify generated content against source evidence and eliminate stale documentation.
Use cases
Use Case: Automated CI/CD Documentation Sync Teams integrate OpenWiki into their CI/CD pipelines to automatically inspect merged PRs and update architectural and API wiki pages against the latest repository state.
Use Case: Measuring Knowledge Grounding and Forgetting AI engineers use the built-in LEDGER benchmark to test how reliably different LLM agent configurations retain facts and avoid hallucinated drift across long series of git commits.
Use Case: Repository Onboarding and Knowledge Scaffolding Developers working with legacy or rapidly expanding codebases run OpenWiki CLI commands to scaffold and maintain grounded technical overviews directly from source code.
Who it is for
OpenWiki is suited for engineering teams, technical leads, and AI researchers looking to automate codebase documentation maintenance or benchmark long-context knowledge grounding. It is not intended for developers seeking an interactive in-editor code completion or general-purpose terminal pair-programming assistant.
Install guide
Installation Install OpenWiki globally or build from source (standard npm package setup): bash npm install -g openwiki # Or clone and build locally: git clone https://github.com/langchain-ai/openwiki.git cd openwiki npm install
First Run Configure your environment with the required LLM API keys: bash export ANTHROPIC_API_KEY="your-anthropic-api-key" export GEMINI_API_KEY="your-gemini-api-key"
Example Session Run an update pass against your current branch to generate or synchronize wiki documentation: bash openwiki update
Flameox — Flameox is an open-source coding agent harness and orchestration framework maintained by morluto. Built to support hybrid development workflows across…
Repo Forensics — Repo Forensics is an open-source agent orchestration harness and repository security gating tool developed by Alex Greenshpan (alexgreensh). Designed to…
Grite — Grite is an open-source terminal coding agent harness developed by Neul Labs. Implemented in Rust, it provides a lightweight command-line…
Agent Memory System — Agent Memory System is an open-source agent orchestration harness developed by RavByte-AI. It provides structured memory retention and retrieval mechanisms…
Mneme — Mneme is an open-source orchestration harness and memory governance framework developed by MnemeHQ designed to prevent architectural drift across the…
GitAgent — GitAgent (formerly gitclaw) is an open-source CLI coding agent developed under the open-gitagent organization and associated with Lyzr. It is…