Fetching from the wire…
Research2026-08-09 · source-backed
arXiv 2608.05604 names the mismatch precisely: current systems retrieve skills as packages but compress them as prose, which destroys the execution contract. SkillZip does contract-preserving compression over section-level graphs, rewriting recurring valid motifs into reversible macros while preserving boundary signatures and dependencies. 3.46x compression, 99.2% dependency preservation, 98.7% verifier reachability, up to 12.2 points of task improvement. A companion pass (ReZip) folds new skills in and revises macros from execution feedback. Demonstrated from 200 to 100,000 skills. Directly relevant if your .claude/skills directory has outgrown its context budget.
Each link below shares sources, entities, or timing with this story.
Introduces temporal causal diagnostics to distinguish legitimate task execution from injected manipulation in multi-turn agent interactions, plus context purification to neutralize poisoned content. Directly applicable to anyone building agents that call external tools. arXiv...
The TypeScript project compiles a repository into an explorable, searchable graph you can ask questions against, with the explicit framing "graphs that teach beat graphs that impress." It targets Claude Code, Codex, Cursor, and Copilot. Directly relevant if your agent currentl...
Addresses the fundamental privacy dilemma: cloud models need data access but enterprises can't share sensitive information. Splits execution between enterprise-side privacy agents and cloud-side capability agents. Directly relevant to AWS AgentCore and enterprise adoption. arX...
Beyond binary human-vs-machine detection: identifies which specific LLM generated a code snippet. Enables vulnerability triage (which model produced the bug?), licensing audits, and distillation detection. Directly relevant to the Anthropic distillation crackdown. (arXiv 2603....
arXiv 2608.00765 compresses retrieved docs into query-conditioned visual representations, sidestepping the trade-off where hard compression is query-aware but weak and soft compression is strong but needs costly offline encoding. Beats both baselines across varying retrieval d...
arXiv 2608.06196 pits lexical+dense ranking against a graph encoding prerequisites, data flow and ordering across 117 realistic non-echoing queries. The ranker hits top-5 in 73.5% ±8.0 of cases; graph neighbours at matched token budget lose 11.2 points at p=0.0007. The mechani...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.