Skip to content
Back to search

CodeNib

Find source context and trace calls across languages in Claude Code, Codex and other MCP agents.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/sysevol-ai/CodeNib with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/sysevol-ai/CodeNib'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/sysevol-ai--codenib/badge.svg)](/listings/sysevol-ai--codenib)

HTML

<a href="/listings/sysevol-ai--codenib"><img src="/api/listings/sysevol-ai--codenib/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 15 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "name": "codenib",
  3. "description": "CodeNib plugins for Claude Code",
  4. "owner": {
  5. "name": "CodeNib Contributors",
  6. "url": "https://codenib.ai"
  7. },
  8. "plugins": [
  9. {
  10. "name": "codenib",
  11. "source": "./plugins/codenib",
  12. "description": "Source-linked answers about your repository: ranked search, compiler-indexed references, and a Wiki whose every claim links to the exact lines."
  13. }
  14. ]
  15. }
README.md · README · 311 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. <!--
  2. SPDX-FileCopyrightText: 2025-2026 CodeNib Contributors
  3. SPDX-License-Identifier: Apache-2.0
  4. -->
  5. <!-- mcp-name: ai.codenib/codenib -->
  6. <div align="center">
  7. <img src="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/codenib_logo.svg" alt="CodeNib" width="440">
  8. <h1>Understand unfamiliar code with evidence.</h1>
  9. <p>
  10. Read a Wiki of how a repository works, follow its call paths on a map,
  11. and open the source behind every explanation.<br>
  12. Then give your coding agent the same index.
  13. </p>
  14. <p>
  15. <a href="https://demo.codenib.ai"><strong>Browse a Wiki</strong></a>
  16. &nbsp;&middot;&nbsp;
  17. <a href="#use-it-with-your-coding-agent">Use with your agent</a>
  18. &nbsp;&middot;&nbsp;
  19. <a href="#hand-your-agent-the-whole-thread">Thread context</a>
  20. &nbsp;&middot;&nbsp;
  21. <a href="https://docs.codenib.ai/">Docs</a>
  22. &nbsp;&middot;&nbsp;
  23. <a href="https://docs.codenib.ai/language_capabilities/">Languages</a>
  24. &nbsp;&middot;&nbsp;
  25. <a href="https://github.com/sysevol-ai/CodeNib/releases">Releases</a>
  26. </p>
  27. <p>
  28. <a href="https://github.com/sysevol-ai/CodeNib/actions/workflows/ci-full.yml"><img src="https://github.com/sysevol-ai/CodeNib/actions/workflows/ci-full.yml/badge.svg" alt="CI"></a>
  29. <a href="https://pypi.org/project/codenib/"><img src="https://img.shields.io/pypi/v/codenib.svg?cacheSeconds=300" alt="PyPI version"></a>
  30. <a href="https://github.com/sysevol-ai/CodeNib/releases/latest"><img src="https://img.shields.io/github/v/release/sysevol-ai/CodeNib" alt="Latest GitHub Release"></a>
  31. <a href="https://arxiv.org/abs/2607.25431"><img src="https://img.shields.io/badge/arXiv-2607.25431-b31b1b.svg" alt="arXiv paper"></a>
  32. <a href="https://huggingface.co/papers/2607.25431"><img src="https://img.shields.io/badge/Hugging_Face-%231_Paper_of_the_Day-FFD21E.svg?logo=huggingface&amp;logoColor=black" alt="Hugging Face: #1 Paper of the Day"></a>
  33. <a href="https://github.com/sysevol-ai/CodeNib/blob/main/LICENSE"><img src="https://img.shields.io/badge/License-Apache_2.0-blue.svg" alt="Apache 2.0 license"></a>
  34. </p>
  35. <p>
  36. <a href="https://docs.codenib.ai/codegraph/#claude-code-plugin"><img src="https://img.shields.io/badge/Claude_Code-plugin-D97757?logo=claude&amp;logoColor=white" alt="Claude Code plugin"></a>
  37. <a href="https://docs.codenib.ai/codegraph/#codex-plugin"><img src="https://img.shields.io/badge/Codex-plugin-10A37F" alt="Codex plugin"></a>
  38. <a href="https://docs.codenib.ai/codegraph/#grok-build-plugin"><img src="https://img.shields.io/badge/Grok_Build-plugin-000000" alt="Grok Build plugin"></a>
  39. <a href="https://docs.codenib.ai/codegraph/#deepseek-harness-bundle"><img src="https://img.shields.io/badge/DeepSeek_Harness-plugin-4D6BFE?logo=deepseek&amp;logoColor=white" alt="DeepSeek Harness plugin"></a>
  40. <a href="https://github.com/marketplace/actions/codenib-wiki"><img src="https://img.shields.io/badge/GitHub_Marketplace-Wiki_action-2088FF?logo=githubactions&amp;logoColor=white" alt="CodeNib Wiki action on GitHub Marketplace"></a>
  41. </p>
  42. </div>
  43. <p align="center">
  44. <a href="https://demo.codenib.ai/psf__requests?p=redirects">
  45. <img src="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/readme/wiki-map-requests.png" alt="The CodeNib Wiki for psf/requests: the page list on the left, a map of the Wiki's areas joined by recorded references in the middle, and the Redirect Resolution page on the right, which traces Session.send() to resolve_redirects() to get_redirect_target() with a file and line for each step" width="100%">
  46. </a>
  47. </p>
  48. <p align="center"><sub>
  49. The <a href="https://demo.codenib.ai/psf__requests?p=redirects">Requests Wiki</a>.
  50. Each node on the map is a page, and the heavy lines are references the code
  51. index recorded between areas. The page cites a file and line range for each claim.
  52. </sub></p>
  53. New in **0.2.6**: Wikis open as a map of their pages and lead with the main path
  54. through the code, one GitHub Action publishes yours to GitHub Pages, plugins
  55. connect Claude Code, Grok Build and DeepSeek Harness, graph builds get faster,
  56. and first C/C++ and JavaScript/TypeScript builds leave the checkout alone. Two
  57. previews serve coding agents: `codenib guard` lists the tests a change can
  58. reach, and [`codenib threads`](#hand-your-agent-the-whole-thread) hands your
  59. agent a GitHub issue or pull request in one read.
  60. [Read the release notes](https://docs.codenib.ai/releases/0.2.6/).
  61. ## A Wiki you can follow down to the source
  62. CodeNib indexes a repository, groups its code into areas, and writes a Wiki
  63. page for each one. The map and the call paths come from the code index, so you
  64. can check every arrow against the source.
  65. - **Start from the map.** Pages are nodes. Drag to pan, scroll to zoom, and
  66. click a node to read its page. Click a heavy line to read an example of the
  67. references behind it.
  68. - **Follow one path through the code.** The Overview opens on the main path
  69. through the repository, one sentence per recorded call. The most important
  70. detail pages draw their longest recorded call path. No model chooses the
  71. arrows.
  72. - **Check each claim.** Sentences end with the file and line range they come
  73. from, and each page lists its callers and calls in the code graph.
  74. - **Ask a follow-up.** *Ask this repository* answers a question with source
  75. citations, using the model you configure.
  76. <p align="center">
  77. <a href="https://demo.codenib.ai/gin-gonic__gin">
  78. <img src="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/readme/wiki-callpath-gin.png" alt="The CodeNib Wiki for gin-gonic/gin: a map of eight areas such as Radix Routing Tree and Engine Lifecycle, beside the Overview, which follows one recorded path from GET() through RouterGroup.handle and Engine.addRoute() to node.addRoute, citing the file and line of each call" width="100%">
  79. </a>
  80. </p>
  81. <p align="center"><sub>
  82. The <a href="https://demo.codenib.ai/gin-gonic__gin">gin Wiki</a> opens on one
  83. recorded path, from <code>GET()</code> to <code>node.insertChild()</code>.
  84. </sub></p>
  85. ### Open a Wiki
  86. | To | Do this |
  87. | --- | --- |
  88. | Read one now | Open a [ready Wiki](https://demo.codenib.ai/browse) for requests, scikit-learn, sympy, fmt, redis, caddy, gin, hugo and more. No account or API key. |
  89. | Map a public GitHub repository | Paste its URL on [demo.codenib.ai](https://demo.codenib.ai), or replace `github.com` with `demo.codenib.ai` in the URL. A new Wiki uses your OpenRouter account, with a $0.25–$5 spending limit you set before it starts. |
  90. | Map your own checkout | Run the Wiki locally, as below. |
  91. | Share it | Publish the Wiki, with its source citations, to GitHub Pages: add the [CodeNib Wiki action](https://github.com/marketplace/actions/codenib-wiki) to a workflow. [Pages guide](https://docs.codenib.ai/github_pages/) |
  92. To build a Wiki with the map on your machine:
  93. ```bash
  94. python -m pip install "codenib[graph]"
  95. codenib wiki /path/to/your/repo --preset graph
  96. ```
  97. This builds a search index and a symbol graph, then opens the Wiki at
  98. `http://localhost:3000`. Pages are written from the index without a model.
  99. Add `--generate --model <provider/model>` for model-written pages.
  100. [Local Wiki guide](https://docs.codenib.ai/quickstart/)
  101. <a id="quickstart"></a>
  102. ## Use it with your coding agent
  103. Build the same index for Claude Code or Codex. Your agent can then find
  104. definitions, follow callers and read the source in your own checkout.
  105. <p align="center">
  106. <picture>
  107. <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/readme/caller-path-dark.svg">
  108. <img src="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/readme/caller-path-light.svg" width="830" alt="The callers of rebuild_auth() in Requests, as one dependency_subgraph call returned them. Eight public entry points, api.request() and Session.get(), options(), head(), post(), put(), patch() and delete(), each reach Session.request() at sessions.py:500. From there one path leads through Session.send() at sessions.py:673 and resolve_redirects() at sessions.py:159 to rebuild_auth() at sessions.py:282.">
  109. </picture>
  110. </p>
  111. One recorded Claude Code session on Requests. Asked where Requests strips
  112. authentication on a redirect to another host, and which public methods reach
  113. that code, the agent found `rebuild_auth()` with `explore_context`. One
  114. `dependency_subgraph` call then returned the path above, each step with its
  115. file and line. The session took 14.1 s, 3 turns and 2 CodeNib tool calls.
  116. [▶ Watch it replay](https://codenib.ai/#get-started)
  117. · [Pinned source, transcript and setup details](https://docs.codenib.ai/codegraph/#recorded-agent-session)
  118. Requires Python 3.10+, Git, and Claude Code or Codex. Initialization accepts
  119. uncommitted edits and preserves the checkout.
  120. ```bash
  121. python -m pip install "codenib[graph,mcp]"
  122. codenib codegraph init /path/to/your/repo
  123. ```
  124. Then ask your agent: **“Use CodeNib's `explore_context` to find where request
  125. retry behavior is implemented. Cite the files and lines.”**
  126. This builds local search and a typed symbol graph, then connects installed
  127. agent clients. CodeNib needs no model, API key, or GPU of its own; your agent
  128. uses its own model. Language toolchain and project prerequisites vary.
  129. In Claude Code you can install the plugin instead. It connects every Git
  130. checkout you open and adds `/codenib:explain`, `/codenib:impact`,
  131. `/codenib:thread`, `/codenib:setup` and `/codenib:wiki`:
  132. ```bash
  133. claude plugin marketplace add sysevol-ai/CodeNib
  134. claude plugin install codenib@codenib
  135. ```
  136. Codex and Grok Build install the same plugin, and DeepSeek Harness installs it
  137. as a bundle:
  138. ```bash
  139. codex plugin marketplace add sysevol-ai/CodeNib --sparse .claude-plugin --sparse plugins/codenib
  140. codex plugin add codenib@codenib
  141. grok plugin install "sysevol-ai/CodeNib#plugins/codenib"
  142. dsh plugin --profile web add "git+https://github.com/sysevol-ai/CodeNib.git#main&path:/plugins/dsh"
  143. ```
  144. [Codex setup](https://docs.codenib.ai/codegraph/#codex-plugin)
  145. · [Grok Build setup](https://docs.codenib.ai/codegraph/#grok-build-plugin)
  146. · [DeepSeek Harness setup](https://docs.codenib.ai/codegraph/#deepseek-harness-bundle)
  147. [Setup and troubleshooting](https://docs.codenib.ai/codegraph/)
  148. · [Source exclusions](https://docs.codenib.ai/codegraph/#select-the-repository-source-surface)
  149. · [All MCP tools](https://docs.codenib.ai/mcp/)
  150. · [Change guard](https://docs.codenib.ai/guard/) (preview, Python: lists the tests a change reaches before your agent finishes)
  151. ## Hand your agent the whole thread
  152. `codenib threads` reads a GitHub issue or pull request once and writes one
  153. document for your agent: the ask, the review comments, the failing checks, the
  154. linked threads, and the code they point at. It runs on your machine and calls
  155. no model. It is a preview in 0.2.6.
  156. <p align="center">
  157. <picture>
  158. <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/readme/thread-tokens-dark.svg">
  159. <img src="https://raw.githubusercontent.com/sysevol-ai/CodeNib/main/assets/readme/thread-tokens-light.svg" width="830" alt="Change in input tokens when the agent is handed the thread context, against reading the thread itself with gh. Codex on 76 pull requests: 21% fewer, with a 95% interval of 27% to 14% fewer. Codex on 47 issues: 25% fewer, 31% to 19%. Claude Code on 76 pull requests: 45% fewer, 57% to 30%.">
  160. </picture>
  161. </p>
  162. Each dot is the change in input tokens over all the tasks when the agent is
  163. handed the prepared context, against reading the thread itself with `gh`, and
  164. each line is its 95% interval. An agent that reads a thread itself makes one
  165. call per piece, and every call sends the conversation so far again; the
  166. prepared document is read once. The agents found the files to change about as
  167. often either way: 0.75 against 0.74 and 0.85 against 0.82 with Codex, 0.80
  168. against 0.83 with Claude Code.
  169. ```bash
  170. cd /path/to/your/repo
  171. codenib threads context 812 # write THREAD_CONTEXT.md before the agent starts
  172. codenib threads open 812 # or start a ticket that remembers what it read
  173. codenib threads inbox 812 # later: only what is new
  174. codenib threads board # all tickets, and other work on the same code
  175. ```
  176. When several agents work on one repository, each ticket is sent only the
  177. records it has not read, and tickets whose work reaches the same function are
  178. told about each other.
  179. [Thread context guide](https://docs.codenib.ai/threads/)
  180. · [How it decides what to include, with real output](https://codenib.ai/blogs/thread-context-for-coding-agents/)
  181. · In Claude Code and Grok Build, the plugin adds it as `/codenib:thread`.
  182. ## How it compares
  183. What an agent can ask each tool. Peer interfaces were checked against upstream
  184. docs on 2026-09-28. This compares capabilities, not benchmark scores.
  185. | Tool | Ask in plain language | Relationships beyond one hop | Relationships resolved by | Runs |
  186. | --- | --- | --- | --- | --- |
  187. | grep / read | No; literal or regex | No | Text match | Local |
  188. | **CodeNib CodeGraph** | Yes; ranked source anchors (`explore_context`) | Incoming/outgoing static references; one `dependency_subgraph` call, depth ≤ 8 | SCIP/LSP indexers (compiler-resolved) | Local |
  189. | [Serena](https://github.com/oraios/serena) | No; symbol name or regex | One level per call (`find_referencing_symbols`) | Live language server or IDE | Local |
  190. | [CodeGraph (colbymchenry)](https://github.com/colbymchenry/codegraph) | Yes; FTS5 full-text (`codegraph_explore`) | Yes; call paths in `codegraph_explore` | Tree-sitter AST extraction | Local; telemetry opt-out |
  191. | [DeepWiki public MCP](https://docs.devin.ai/work-with-devin/deepwiki-mcp) | Yes; generated answers (`ask_question`) | No structured graph tools | Generated Wiki | Hosted |
  192. CodeNib's indexed references can include calls, fields, type uses and function
  193. values. Inspect the source to establish which references invoke a function;
  194. external and dynamic relationships may be missing. For a known string or file,
  195. your agent's native search and reads remain useful. Include indexing and rebuild
  196. costs when assessing graph navigation on your own tasks.
  197. None of these repository tools needs its own model or API key; your agent
  198. uses its own model. [Detailed comparison, sources and boundaries](https://docs.codenib.ai/comparison/).
  199. CodeNib also includes an optional [grep → Jev route](https://docs.codenib.ai/guides/grep-jev/)
  200. using OpenRouter for planning and reranking; selected code goes to remote
  201. models. Authorization previews remain opt-in.
  202. ## Languages
  203. CodeNib registers 14 languages, 12 of them with a code graph.
  204. | Support | Languages |
  205. | --- | --- |
  206. | Code graph with an incremental backend (LSP or clangd) | Python, Go, Rust, C/C++, JavaScript, TypeScript |
  207. | Code graph | C#, Java, Ruby, PHP, Kotlin, Scala |
  208. | Search only, no definitions or callers | Swift, Lua |
  209. These are registered capabilities, not the tools on your machine;
  210. `codenib toolchain status` shows those. The product currently reuses or
  211. rebuilds views; file-level delta repair is not enabled. The
  212. [language matrix](https://docs.codenib.ai/language_capabilities/) lists each
  213. backend and its prerequisites.
  214. ## Results and reproductions
  215. - **#1 in line recall and Pass@1 on ContextBench's Agent systems leaderboard**:
  216. 77.90% line recall and 388 of 500 tasks resolved (77.60% Pass@1), as of
  217. October 23, 2026 (UTC). This is our research agent, not the Wiki or the code
  218. graph: two Claude Haiku 5.5 workers search with glob, grep and reads, then
  219. Claude Opus 5.5 repairs.
  220. [How it works](https://codenib.ai/blogs/contextbench-500-tasks/)
  221. · [Scope and evidence](https://codenib.ai/blogs/contextbench-500-tasks/#scope)
  222. - **71.4% code-block Recall@5 on 100 real issues** with model-planned
  223. grep → Jev, 12.8 percentage points above the same grep candidates without
  224. reranking. This is an experimental retrieval route, separate from the
  225. model-free CodeGraph.
  226. [Method and limitations](https://docs.codenib.ai/evaluation/grep_jev/)
  227. | Start here | What you can inspect |
  228. | --- | --- |
  229. | [Context methods and support matrix](https://docs.codenib.ai/concepts/methods/) | Every retrieval pipeline, LLM workflow and agent route, sorted by the views it reads and who controls the search |
  230. | [Benchmarks](https://docs.codenib.ai/evaluation/) | Localization, agent-context and resolution layers, evidence levels, and the method × benchmark matrix over CodeNib Base/Synthesis, SWE-bench Lite, Loc-Bench, SWE-Explore, Agent Retrieval Bench and ContextBench |
  231. | [Agent integration matrix](https://docs.codenib.ai/agent_integrations/) | Revision-pinned LocAgent, Agentless, CoSIL, OrcaLoca and RepoNavigator contracts; compatibility does not imply reproduced paper scores |
  232. | [SWE-Explore validation](https://docs.codenib.ai/evaluation/swe_explore/) | 1,020/1,020 real-output metric cells match the pinned official evaluator on a fixed 20-case run |
  233. | [DGX Spark deployment](https://docs.codenib.ai/guides/reference-deployments/dgx-spark/) | Local Wiki, CodeGraph and model serving on GB10; a deployment guide, not a token-saving benchmark |
  234. ## Documentation and community
  235. [Documentation](https://docs.codenib.ai/)
  236. · [Architecture](https://docs.codenib.ai/concepts/architecture/)
  237. · [Contributing](https://github.com/sysevol-ai/CodeNib/blob/main/CONTRIBUTING.md)
  238. · [CI and testing](https://docs.codenib.ai/ci_cd/)
  239. · [Changelog](https://github.com/sysevol-ai/CodeNib/blob/main/CHANGELOG.md)
  240. · [Discord](https://discord.com/invite/AtrukE7Jpx)
  241. CodeNib is in beta; public interfaces may change before a stable release.
  242. Historical research artifacts retain their published dataset identifiers.
  243. ## Citation
  244. If you use CodeNib in your research, please cite our
  245. [arXiv paper](https://arxiv.org/abs/2607.25431):
  246. ```bibtex
  247. @misc{yu2026codenibmultiviewdataserving,
  248. title={CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents},
  249. author={Zhongming Yu and Hengjia Yu and Boqin Yuan and Shuting Zhao and Yizhao Chen and Aryan Dokania and Mihir Jagtap and Jiayu Chang and Yitong Ma and Yash Jayswal and Wentao Ni and Hejia Zhang and Zhaoling Chen and Gangda Deng and Jishen Zhao},
  250. year={2026},
  251. eprint={2607.25431},
  252. archivePrefix={arXiv},
  253. primaryClass={cs.SE},
  254. url={https://arxiv.org/abs/2607.25431},
  255. }
  256. ```
  257. CodeNib is licensed under [Apache 2.0](https://github.com/sysevol-ai/CodeNib/blob/main/LICENSE).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.