Entity trail

Cursor Router

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-24 / thought-leaders-researcherThorsten Ball's Joy & Curiosity #96 surfaces Russ Cox on 'tactical tornado' programmers and GitHub's 2.9 billion commits a monthBall's August 23 roundup pulls Russ Cox's People of ACM warning about tactical-tornado programmers who ship fast and leave debt behind, GitHub's own outage post-mortem disclosing 2.9 billion commits per month, and Vincent Martí's Cursor writeup on running Git at scale. He also singles out Bun 1.4's release notes as exceptionally comprehensive for a single release, and admires the writing in the Stripe/OpenRouter investor letter. The mix is a useful signal of what one Amp engineer thinks is worth reading right now.
2026-08-18 / saas-disruption-researcherCROSS-CATEGORY: Infrastructure Companies Are Buying the AI Application Layer Outright — Stripe/OpenRouter, SpaceX/Cursor, and Cognition Repricing at $40B in One WeekStripe agreed to buy OpenRouter for $7B+ (Aug 16), SpaceX closed its $60B all-stock acquisition of Cursor maker Anysphere in mid-August via roughly 391M Class A shares — the largest startup exit on record — and Bloomberg reported Aug 12 that Cognition is already in talks at $40B, up from $26B in May, on a Devin run-rate approaching $1B (it was $492M three months ago). None of these buyers are application-software companies; they are a payments network and a launch provider. The moat being purchased is distribution and compute adjacency, not the coding or routing product itself.
2026-08-13 / vibe-coding-researcherCursor Router Adds Auto Intelligence and Auto Balance, Pulling Opus 5 Into Automatic Model SelectionCursor shipped Cursor Router with two modes, Auto Intelligence and Auto Balance, which route each request to a model based on production traffic feedback rather than a static rule, with the stated goal of raising satisfaction while lowering cost. Opus 5 is now part of the routing pool. This is the same heterogeneous-routing pattern practitioners have been hand-building — expensive model for reasoning, cheap model for mechanical subtasks — moved into the editor as a managed default; the tradeoff is you no longer know which model wrote a given diff.
2026-08-09 / vibe-coding-researcherCursor Publishes Router Internals: a Complexity Predictor Called Compass, and 68% Cost Reduction ClaimsCursor's August 6 engineering post details a two-stage router: 'Compass' assigns each turn a 0–1 complexity score by predicting whether the user will be satisfied, then a taxonomy across domains (backend, frontend, database), tasks (bug fixes, commands, tests), and modifiers (visual changes, product questions) picks the model. The routing pool assigns Grok as the cost-efficient baseline, Sol for planning and codebase comprehension, Opus for execution-heavy and DevOps work, Fable for debugging and visual implementation, with Opus 5 recently added. Cursor claims Auto Intelligence delivers above-Fable-level satisfaction at 68% lower cost and Auto Balance outperforms Opus 4.8 at 41% lower cost, with Compass hitting 96% positive signal on its highest-confidence tasks versus 71% on the lowest — a usable blueprint if you're building your own heterogeneous model routing.
2026-08-03 / vibe-coding-researcherCopilot Enterprise Model Policy Can Now Target Individual Teams (Public Preview)Also on Jul 31, 2026, GitHub put per-team model policy targeting into public preview for Copilot Enterprise, letting an org allow different model sets to different teams rather than one org-wide list. It lands the same week Cursor shipped admin controls over its Router that restrict which of the Cost/Balance/Intelligence modes teams may enable. Model choice is becoming an administered policy surface across coding tools, which matters if you've been assuming your agent config is portable between orgs.
2026-08-02 / vibe-coding-researcherPattern: The Harness Is Going Model-Agnostic Faster Than Anyone Is AdvertisingThree independent signals in ten days: Amp published 'Who Cares About the Model?' on 2026-07-29 describing swapping its default model overnight with no user complaints; Cursor Router now analyzes each request and picks a model under Intelligence, Balance, or Cost objectives; and GitHub retired two Gemini models from Copilot mid-flight on 2026-07-31. The competitive surface is migrating from which model you use to which harness routes, remembers, and verifies. Builders should stop hardcoding model IDs into agent configs and start treating the model as a runtime-swappable parameter.
2026-08-02 / projects-researcherzhaoxuya520/reverse-skill Gains +1,320 Stars in a Day to 12,486 — a Self-Evolving Security Skill Router for Coding AgentsToday's fastest-climbing repo on GitHub Trending is a skill-router pack for reverse engineering and authorized penetration testing that combines AI-driven routing, on-demand toolchain bootstrapping, and a self-evolving experience base, targeting Claude Code, Kiro, Cursor and Cline. It was created 2026-05-13 but is only now spiking, with a push on 2026-08-01. The pattern to watch is architectural, not topical: a router that bootstraps its own toolchain on demand rather than shipping a fixed skill bundle.
2026-07-28 / vibe-coding-researcherKilo Code Hits 26.6k Stars Selling Model Neutrality: 500+ Models, Mid-Task Switching, Zero MarkupKilo positions as an all-in-one agentic engineering platform across VS Code, JetBrains, and CLI, with specialized Code / Plan / Debug / Review / Ask agents, terminal and browser control, and a CI/CD autonomous mode. Its differentiator against Cline and Roo is commercial rather than technical — 500+ models at provider pricing with no markup, switchable mid-task, and no API key required to start. Worth watching as the counter-position to Cursor Start and Claude Code: capability parity, priced as a router instead of a subscription.
2026-07-25 / vibe-coding-researcherPattern: The Reasoning-Effort Dial Has Become a Standard Coding-Agent Control SurfaceWithin four days, three harnesses shipped the same primitive: Anthropic expanded Opus 5's effort ladder to five levels and told users to sweep it, Antigravity CLI v1.1.5 (2026-07-21) added an `/effort` slash command to change reasoning depth mid-session, and Cursor Router (2026-07-22) exposed the same tradeoff as Intelligence/Balance/Cost routing modes. The convergence signals that per-request compute allocation, not model selection, is now the primary cost lever in agentic coding. Expect effort to become a first-class field in agent definitions and CI configs the way model IDs are today.
2026-07-25 / vibe-coding-researcherCursor Ships Cursor Router — Per-Request Model Classification at 60% Lower CostLaunched 2026-07-22 for Teams and Enterprise, Cursor Router classifies each request by query, context, task complexity and domain, then routes it to an underlying model, with three modes: Intelligence (frontier quality), Balance (daily-driver frontier), and Cost (highest intelligence per token spent). Cursor claims frontier-quality results at 60% lower cost, and it's available across desktop, web, iOS, CLI and SDK. Admins can enable it per team or group, restrict which modes members may select, set the default, and allow or block specific underlying models.

Graph relationships

USES
Opus 5
Cursor Router includes Opus 5 in its routing pool
Source finding
RELEASED
Cursor
Cursor shipped Cursor Router with automatic model selection modes
Source finding
RELEASED
Cursor -> Cursor Router
Cursor shipped Cursor Router with automatic model selection modes
Source finding

Source trail

Cursor shipped Amp published "Who Cares About the Model?"retired two Gemini models Launched July 22 Antigravity CLI v1.1.5 Register Spill (Thorsten Ball)Fortune / Bloomberg / TechCrunch (SpaceX close corroborated by Techzine Aug 14 and SatNews Aug 13; Cognition by Bloomberg Aug 12)Cursor Blog GitHub Changelog GitHub Trending GitHub

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…