CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents

Hugging Face Daily Papers Papers
coding-agents repository-context multi-view data-system search navigation

Summary

CodeNib is a multi-view data system that serves repository context to coding agents by building reusable lexical, dense, and structural views per commit, enabling faster updates and efficient context serving.

Coding agents repeatedly search, navigate, and retain context from evolving repositories, but disconnected indexes, language servers, and task-local histories force repeated discovery and obscure lifecycle costs. CodeNib builds reusable lexical, dense, and structural views per repository commit, maps outputs to repository-relative source ranges, maintains selected views across edits, and serves ranked search, symbol navigation, and bounded context through one runtime. Across 100 snapshots, we map quality-cost frontiers across the repository-context lifecycle. When outputs match an independent rebuild, graph and vector updates are 8.7times and 25.4times faster at the median. On the static-navigation subset matching normalized live-server locations (63% of 1,000 requests), the median per-request live/static latency ratio is 4.7times. Across five models, selected context policies preserve localization with 50--87% fewer trajectory tokens than paired grep/read. Together, these results support multi-view repository-context serving with explicit, operation-specific validity boundaries.
Original Article
View Cached Full Text

Cached at: 07/29/26, 11:52 AM

Paper page - CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents

Source: https://huggingface.co/papers/2607.25431 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Codingagentsrepeatedlysearch,navigate,andretaincontextfromevolvingrepositories,butdisconnectedindexes,languageservers,andtask-localhistoriesforcerepeateddiscoveryandobscurelifecyclecosts.CodeNibbuildsreusablelexical,dense,andstructuralviewsperrepositorycommit,mapsoutputstorepository-relativesourceranges,maintainsselectedviewsacrossedits,andservesrankedsearch,symbolnavigation,andboundedcontextthroughoneruntime.Across100snapshots,wemapquality-costfrontiersacrosstherepository-contextlifecycle.Whenoutputsmatchanindependentrebuild,graphandvectorupdatesare8.7timesand25.4timesfasteratthemedian.Onthestatic-navigationsubsetmatchingnormalizedlive-serverlocations(63%of1,000requests),themedianper-requestlive/staticlatencyratiois4.7times.Acrossfivemodels,selectedcontextpoliciespreservelocalizationwith50--87%fewertrajectorytokensthanpairedgrep/read.Together,theseresultssupportmulti-viewrepository-contextservingwithexplicit,operation-specificvalidityboundaries.

View arXiv pageView PDFProject pageGitHub7Add to collection

Get this paper in your agent:

hf papers read 2607\.25431

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.25431 in a model README.md to link it from this page.

Datasets citing this paper1

#### sysevol-ai/codenib-synthesis Viewer• Updatedabout 2 hours ago • 500 • 248 • 2

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.25431 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

@appliedcompute: https://x.com/appliedcompute/status/2052826576723841292

X AI KOLs Timeline

Applied Compute introduces ACL-Wiki, a continual learning memory system built on their Context Engine that logs coding agent interactions from Cursor, Claude Code, and Codex to build an improving Contextbase, roughly doubling the Critical Memory Rate over two weeks. The system uses a Remember-Refine-Retrieve pipeline exposed via MCP server to give coding agents institutional memory that improves with use.