repoGitHubTrust 82 · PrimaryPublished yesterdayLive · 14h ago
dean0x/skim
The most intelligent context optimization engine for coding agents. Code-aware AST parsing across 17 languages. Command rewriting. Test, build, and git output compression. Token budget cascading. Built in Rust.
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 56%Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models →
- PossiblePossibly related (embedding) · 55%ACE: Pluggable Adaptive Context Elasticizer across Agents →
- PossiblePossibly related (embedding) · 55%A Sovereign, Open-Source Foundation Model for German and English →
- PossiblePossibly related (embedding) · 54%RuBench: A Repository-Level Agentic Coding Benchmark with Natively Authored Russian Task Specifications →
- PossiblePossibly related (embedding) · 54%PolyWorkBench: Benchmarking Multilingual Long-Horizon LLM Agents →
Implements
paperFunction-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation ModelspaperACE: Pluggable Adaptive Context Elasticizer across AgentspaperA Sovereign, Open-Source Foundation Model for German and EnglishpaperRuBench: A Repository-Level Agentic Coding Benchmark with Natively Authored Russian Task SpecificationspaperPolyWorkBench: Benchmarking Multilingual Long-Horizon LLM Agents
Related across the graph
paperPolyWorkBench: Benchmarking Multilingual Long-Horizon LLM AgentspaperACE: Pluggable Adaptive Context Elasticizer across AgentspaperA Sovereign, Open-Source Foundation Model for German and EnglishpaperRuBench: A Repository-Level Agentic Coding Benchmark with Natively Authored Russian Task SpecificationspaperFunction-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models