groundy

agents & frameworks

102 articles·rss

Top in agents & frameworks

agents

Multi-Agent or Single-Agent LLM: What Skill Distillation Actually Costs

arXiv 2604.01608 maps when collapsing multi-agent teams into one model saves cost. Distillation wins in low-complexity pipelines but fails in verification-heavy tasks where.

aug 20
agents

Multi-Agent LLM Systems Drift Into Misaligned Communication Over Long Horizons

A new arXiv study found 12.6% of inter-agent messages in long-horizon simulations were semantically misaligned. Protocol compliance does not prevent this drift, so teams must.

aug 20
agents

Sizing Agent Memory: A Capacity Planning Rubric for Long-Horizon LLMs

Measure input tokens, gate stage boundaries, and size backbones before adding retrieval. A capacity planning rubric for agent memory based on August 2026 arXiv data.

agents

When Spec-First Agents Dismantle Invariants: A Governance Case Study

A specification-first coding agent eroded a core invariant across 189 files without review. The fix is executable conformance checks, not more human oversight.

agents

Cloudflare Kitesurf: V8 Isolates vs Containers for Agent Browsing

Cloudflare Kitesurf uses V8 isolates for agent browsing instead of containers. This shifts costs from memory to rendering fidelity. Compare edge isolates with Playwright.

agents

Claude Code Skills vs Model Weights: Where Should Agent Skills Live?

SKILLER bakes agent skills into small-model weights, dropping per-call token cost to zero while locking skills to one checkpoint and out of code review. Route by frequency.

agents

Cloudflare Precursor: Behavioral Detection for AI Agents

Cloudflare Precursor reportedly shifts AI agent detection from spoofable headers to continuous behavioral signals. This article verifies the claims and outlines operator.

agents

Why Multi-Agent LLM Delegation Concentrates Risk

Splitting tasks across planner and worker agents concentrates risk because principals cannot observe hidden actions. Base models defect. Teams must add verifiable execution.

more in this beat1–50 of 94 · page 1 / 2
  1. jul 29agentsHarness vs Scaffold: Why Claude Code and LangGraph Are Not Interchangeable
  2. jul 28agentsMCP Tool Discovery Moves From Hardcoded Config to Runtime Agent Search
  3. jul 28agentsCode-as-Action Agents Beat GAIA But Require Runtime Sandboxing
  4. jul 27agentsWhy Agent Security Tests Must Audit Full Trajectories, Not Single Turns
  5. jul 26agentsx402 Per-Call Payments: Agent Wallet Custody and Replay Risks
  6. jul 25agentsCodeRabbit Review Study: 56% Rejection Rate Demands Targeted Scoping
  7. jul 24agentsLLM Agents Ignore Mid-Flight Halt Signals: 0 of 40 Trials Stopped
  8. jul 24agentsAgent-First CLIs: Why GitHub, npm, and PyPI Must Publish Machine-Readable Contracts
  9. jul 22agentsMCP and AGENTS.md Standardize Context, Not Agent Coordination
  10. jul 22agentsCursor's Swarm Math: When Cheap Agents Save Money and When They Fail
  11. jul 21agentsRuntime monitoring beats alignment for agent-to-agent coercion
  12. jul 20agentsCloudflare's Agent Stack: Edge Trust, Identity, and Metering
  13. jul 19agentsLM Studio Bionic vs Claude Code: Local-First vs Cloud Agent Tradeoffs
  14. jul 17agentsCan a Malicious AGENTS.md File Compromise Your Coding Agent? A Threat Model
  15. jul 17agentsHow GitHub Projects Actually Adopt Coding Agents: New Empirical Data
  16. jul 14agentsWhy CLI Coding Agents Derail Mid-Run, Not at the First Mistake
  17. jul 13agentsTest-Time Scaling Cost Falls as PRMs Reuse Generator KV-Cache
  18. jul 13agentsClaude Code Skills vs Cursor Rules vs MCP: How Agent Skill Systems Compare
  19. jul 12agentsTTHE: Test-Time Harness Evolution Changes the Test-Code Contract for Coding Agents
  20. jul 12agentsGit-for-Data for Agentic Lakehouses: Why Agents Need Versioned State
  21. jul 11agentsGame Theory Can Cut Multi-Agent LLM Hallucination, But Only If Payoffs Align
  22. jul 11agentsWebSwarm: Recursive Multi-Agent Search vs Flat Orchestration
  23. jul 10agentsDo Coding Agents Memorize Their Benchmarks? DeepSWE Tests on Unseen Tasks
  24. jul 10agentsCan a 4B Model Run a Coding Agent? Terminus-4B vs Claude and GPT-4o
  25. jul 09agentsCan You Prove a Governed AI Agent Actually Ran the Action You Authorized?
  26. jul 09agentsAgentTether Repairs LLM Agent Failures with a Runtime Graph
  27. jul 09agentsCan Multi-Agent LLM Negotiation Protocols Trust Their Own Samplers?
  28. jul 09agentsWhy Your AI Travel Agent Would Book a Bullfight
  29. jul 08agentsCan Multi-Agent RAG Run Air-Gapped? A Forensics System Shows How
  30. jul 08agentsCan You Prove an Agentic Trading Pipeline Has No Look-Ahead Bias?
  31. jul 08agentsHow Far Ahead Can a Coding Agent Plan? The Horizon Bottleneck
  32. jul 07agentsSymbolic Inference Forces Agent Frameworks to Expose Intermediate State
  33. jul 06agentsBOUNDARY_SYNC: Why Multi-Agent Representational Coupling Is the New Coordination Failure Mode
  34. jun 30agentsDo Multi-Agent RAG Systems Write Better READMEs Than One Agent?
  35. jun 29agentsAgentic AI Turns Location Trails Into a Re-Identification Tool
  36. jun 29agentsHow a Human-Agent Team Lifts One Video Into 4D Interactions
  37. jun 29agentsCan LLM Agents Learn Cooperation Laws From Embodied Play?
  38. jun 29agentsGovern the Repo, Not the Agent: A New Risk Metric for AI-Native Code
  39. jun 28agentsCan an AI Agent Catch Cryptographic Misuse Before It Ships? Chai Tests the Claim
  40. jun 27agentsCan Spec-Driven Development Keep AI Coding Agents From Drifting?
  41. jun 26agentsCan Knowledge-Based Pull Requests Make Agent Contributions Auditable?
  42. jun 26agentsDo AI Agents Hold Up Outside Familiar Environments? A New Eval Says No
  43. jun 26agentsHow Much Repo Structure Does a Coding Agent Actually Need?
  44. jun 26agentsMCP vs A2A: Two Agent Protocols, One Integration Layer Decision
  45. jun 25agentsCan You Rewind an AI Agent Mid-Run? Reversible Traces Say Yes
  46. jun 25agentsCan AI Agents Reproduce Published Research? CORE-Bench Tests It
  47. jun 25agentsHow On-Device AI Agents Keep Learning by Forgetting on Purpose
  48. jun 25agentsDo AGENTS.md Files Actually Help Coding Agents? A New Benchmark Tests It
  49. jun 25agentsShould AI Shopping Agents Pay Micro-Transactions for Verified Product Data?
  50. jun 24agentsCan a Conversational Graph Compile Into a Goal-Oriented Dialogue Runtime?
about this beateditorial framing

Agent frameworks ship faster than the rigor operators need to run them. Vendor docs promise orchestration, memory, and tool use; academic benchmarks and production post-mortems keep exposing the same structural gaps: diversity collapse in multi-agent ideation, hallucination amplification across consensus topologies, missing per-step rationale traces, role-based retry losing to graph-state failure isolation on long tasks, and configuration surfaces that punish static templates. This beat covers that delta.

The second through-line is governance and trust. Skill registries, tool-use protocols, and capability manifests are accumulating faster than auditable contracts for them. Trust schemas, contractual skill specs, and information-flow controls are arriving as bolt-ons rather than primitives, while the infrastructure layer — sandbox execution, private networking, agent memory — keeps absorbing functionality the framework layer used to own. The question of where the agent stack actually lives, and who is liable when it misbehaves, stays unresolved.

Coverage is comparative and opinionated. When a benchmark or paper exposes a gap that a major framework cannot close without redesign, that gets named. When a vendor ships governance theater rather than enforcement, that gets named too. The goal is help readers pick stacks that survive contact with production, not a taxonomy of every framework release.