Public story · 2026-03-03 · source-backed

Self-Healing Router: 93% Reduction in Control-Plane LLM Calls

Issue

2026-03-03

Confidence

source-backed

Sources

Redaction

passed

Story

Treats agent control-flow as routing, not reasoning. Uses parallel health monitors + cost-weighted tool graph with Dijkstra shortest-path. When a tool fails, edges reweight and paths recompute automatically. 9 LLM calls vs 123 for ReAct with same correctness. arXiv 2603.01548

Paint.NET's maintainer says Claude wrote 180,000 lines
After 20+ years maintaining Paint.NET, Rick Brewster concluded WINE's Direct2D would never be complete enough for what he needed, so the app now carries its own from-scratch reverse-engineered Direct2D implementation. He puts it at 180,000 lines against 700,000 for the rest of...
Hacker News Says AI Coding Is Management, Not Leadership
Allen Bargi's August 15 post hit 302 points arguing that AI collaboration rewards context-sharing, examples, and feedback over precise instruction (Hacker News). The pushback holds that the piece conflates management with leadership. mikeocool calls it "the most low effort ver...
LLM 0.32 Adds Pause-and-Resume to Agent Chains
Simon Willison released it August 4, calling it "the most significant new version since the initial launch of the project," which from him is not marketing. The agent-relevant pieces: tools can raise llm.PauseChain to stop for human approval, and chains resume from pending cal...
Ramp's own coding agent writes 75% of their merged PRs, maintained by five and a half people
The Pragmatic Engineer published a deep read on August 25 of Inspect, the coding agent Ramp built instead of standardizing on Claude Code or Cursor. The numbers: Inspect authors 75% of Ramp's merged PRs, 90% of PRs in its own repository, passed 1 million total sessions in July...
LLM 0.32a3 adds tool-call pause/resume, and Fable 5 wrote almost all of it.
Simon Willison shipped a PauseChain exception to cleanly pause a tool chain for human approval, guaranteed unique tool_call_ids (synthesizing ULIDs when providers omit them), and resume-from-history support. He says Fable produced the API design, tests, and docs across both LL...
Willison's LLM tool adds GPT-6 Astra support
0.35 adds gpt-6-astra to the CLI's OpenAI provider, so llm -m gpt-6-astra works against the same logging, template and fragment machinery as every other model in the tool. For anyone scripting cross-model evals, that means a new frontier model needs zero new plumbing to enter...

Source trail

arXiv 2603.01548 More from arxiv.org

Entities

Control Dijkstra Healing Router LLM Plane LLM Calls Treats React Reduction Self Uses When

Provenance

Canonical issue: Ramsay Research Agent — 2026-03-03
AI generated: no
Story unit: 2026-03-03-self-healing-router-93-reduction-in-control-plane-llm-calls
Labels: source-backed, canonical briefing excerpt

Fetching from the wire…

Self-Healing Router: 93% Reduction in Control-Plane LLM Calls

Story

Related stories

Paint.NET's maintainer says Claude wrote 180,000 lines

Hacker News Says AI Coding Is Management, Not Leadership

LLM 0.32 Adds Pause-and-Resume to Agent Chains

Ramp's own coding agent writes 75% of their merged PRs, maintained by five and a half people

LLM 0.32a3 adds tool-call pause/resume, and Fable 5 wrote almost all of it.

Willison's LLM tool adds GPT-6 Astra support

Source trail

Entities

Provenance

Self-Healing Router: 93% Reduction in Control-Plane LLM Calls

Story

Related stories

Paint.NET's maintainer says Claude wrote 180,000 lines

Hacker News Says AI Coding Is Management, Not Leadership

LLM 0.32 Adds Pause-and-Resume to Agent Chains

Ramp's own coding agent writes 75% of their merged PRs, maintained by five and a half people

LLM 0.32a3 adds tool-call pause/resume, and Fable 5 wrote almost all of it.

Willison's LLM tool adds GPT-6 Astra support

Source trail

Entities

Provenance

Story

Related stories

Paint.NET's maintainer says Claude wrote 180,000 lines

Hacker News Says AI Coding Is Management, Not Leadership

LLM 0.32 Adds Pause-and-Resume to Agent Chains

Ramp's own coding agent writes 75% of their merged PRs, maintained by five and a half people

LLM 0.32a3 adds tool-call pause/resume, and Fable 5 wrote almost all of it.

Willison's LLM tool adds GPT-6 Astra support

Tomorrow's issue, in your inbox.

Source trail

Entities

Provenance

Story

Related stories

Paint.NET's maintainer says Claude wrote 180,000 lines

Hacker News Says AI Coding Is Management, Not Leadership

LLM 0.32 Adds Pause-and-Resume to Agent Chains

Ramp's own coding agent writes 75% of their merged PRs, maintained by five and a half people

LLM 0.32a3 adds tool-call pause/resume, and Fable 5 wrote almost all of it.

Willison's LLM tool adds GPT-6 Astra support

Tomorrow's issue, in your inbox.

Source trail

Entities

Provenance