Blog
Notes on autonomous coding agents, shipping software from issue to PR, and how we build Sinatra.
GitHub Copilot coding agent pricing: what the 2026 AI credits change actually costs
GitHub moved Copilot to usage-based AI credits on June 1, 2026. What the coding agent costs per session now, what each plan includes, and where Actions minutes come in.
A coding agent opened 306 PRs on our repos since April. 289 merged.
Nineteen weeks of real data from running a background coding agent on two production codebases: the merge rate, what got closed without merging, and why the human gate is doing most of the work.
Self-hosted AI coding agents (2026): OpenHands, Centaur, Cyrus
Self-hosted coding agents that turn tickets into PRs: OpenHands, Centaur, Cyrus, Claude Code on your own runners, and what running one actually costs you.
AI coding agent cost for a 10-person startup, in real numbers
Whole-number math for a small team: seats, model bills, and what changes when everyone already pays for Claude or ChatGPT. Plus 50% off for startups.
How to use Claude Code with Linear (three ways, 2026)
Claude Code has no assign-to-Claude in Linear. Three ways to connect them: an MCP connector, the GitHub Action, or a wrapper that assigns from the ticket.
Linear coding sessions explained: pricing, setup, and limits
Linear runs coding sessions now: delegate an issue, get a PR, pay tokens at provider rates plus $0.25 per 20-minute block. How it works and what a month costs.
AI coding agent pricing compared (2026): what 12 agents charge
AI coding agent pricing in 2026, side by side: what 12 vendors charge, what each one meters, and whether your own key or subscription counts in the cloud.
OpenAI Codex cloud pricing explained: a rate limit, not a rate
What Codex cloud costs on each ChatGPT plan, why OpenAI publishes a message range instead of a price, and what the API-key path gives up.
Cursor Cloud Agents, formerly Background Agents: what they cost
Cursor renamed Background Agents to Cloud Agents. How you start one, what a month costs on each plan, the Teams token rate on your own key, and where it stops.
Google Jules review (2026): the free tier is the product
A Google Jules review for engineers: the label-to-PR loop, the 15-task free tier, what Gemini-only and no Linear trigger mean, and who should pick it.
Codegen alternatives: how to keep turning Linear tickets into PRs
Codegen shut down after the ClickUp acquisition. Honest Codegen alternatives for the Linear ticket to PR workflow, plus a migration checklist.
The best coding agents for Linear, compared (2026)
The coding agents for Linear compared for 2026: how you trigger each, where the code runs, what comes back, and what every option costs.
Copilot coding agent vs Jules vs Codex (2026)
Copilot coding agent vs Jules vs Codex: how the GitHub, Google, and OpenAI background agents compare on triggers, models, pricing, and lock-in.
OpenHands review (2026): the open-source coding agent
An OpenHands review for engineers: what the open-source agent platform does well, what self-hosting actually costs, and when its cloud makes sense.
The best model for coding agents depends on the ticket
There is no single best model for coding agents. What differs between frontier models on real tickets, and how to choose per repo without benchmarks.
Devin vs GitHub Copilot coding agent: which fits your team?
Devin vs GitHub Copilot coding agent, compared for 2026: triggers, where code runs, model lock-in, pricing, and enterprise posture.
Devin review (2026): strengths, struggles, and pricing
A fair Devin review for 2026: what Cognition's autonomous engineer does well, where it struggles, current Devin pricing, and who it actually fits.
GitHub issue to pull request automation: every option in 2026
Every route from GitHub issue to pull request automation in 2026: Actions workflows, Copilot coding agent, OpenHands, Devin, and Sinatra, compared.
The best background coding agents in 2026, compared
The best background coding agents in 2026, compared: Sinatra, Devin, Copilot coding agent, Jules, Codex cloud, OpenHands, Factory, Tembo, and Cursor.
How we sandbox AI coding agents
Sandboxing AI coding agents against private code: the threat model, ephemeral sandboxes, audited egress, scoped tokens, and the PR as the boundary.
Why an AI agent should fix its own failing CI
A PR with red checks is unfinished work. What it takes for an AI agent to fix failing CI: read the actual failure, patch it, and push to its own branch.
How to automatically fix GitHub issues with an AI agent
The label-driven workflow for letting an AI fix GitHub issues: how it runs, which issues delegate well, and the guardrails that keep you in control.
Turning Linear tickets into GitHub PRs with an autonomous agent
How Sinatra takes a Linear or GitHub issue and hands back a reviewed pull request: the workflow, what happens in the sandbox, and where you stay in control.
Are AI coding agents safe for private repos? 8 questions
Are AI coding agents safe to connect to a private repo? Eight questions to ask any vendor, why each matters, and what a good answer sounds like.
BYOK for AI coding agents: what bring your own key actually costs
The three ways coding agent vendors charge, what BYOK gets you with AI coding agents, and the honest tradeoffs of holding your own key.
How to assign a Linear issue to an AI agent (and get a PR back)
How to assign a Linear issue to an AI agent and get a pull request back: setup, writing the ticket, reviewing the PR, and requesting changes.
AI backlog grooming, except the tickets get done
Every backlog has a tail of tickets that will never be scheduled. AI backlog grooming works better when an agent ships fixes instead of sorting the pile.
How to review AI-generated PRs
Reviewing AI-generated PRs is a different job than reviewing a teammate's. A checklist in priority order, plus what to do when the diff is wrong.
How to write tickets for AI agents
How to write tickets for AI agents: behavior-first titles, real acceptance criteria, constraints worth stating, and a copyable template.
What is a background coding agent?
A background coding agent works away from your editor: triggered from an issue, run in a sandbox, delivered as a pull request. Definition and checklist.