Resources
Answers to the questions engineering teams ask when they run coding agents, plus step-by-step guides for doing that work in Capy.
Coding agents
How coding agents take on work, run in parallel or in the background, and bring it back for review.
Coding agentsGetting coding agents to verify browser behavior
How coding agents test a change in a real browser and in Safari on a cloud iPhone: what that catches that tests miss, what the agent needs, and an example.
Coding agentsRunning parallel coding agents without shared-tree collisions
What collides when parallel coding agents share one machine, which work can share and which needs its own machine, and a test for choosing.
Coding agentsOrchestrating coding agents across a team
How to orchestrate coding agents for a whole team: one intake, one set of rules, an owner for every decision, and one coordinator per area.
Coding agentsWhat is an agent swarm for engineering?
An agent swarm is a group of coding agents working one goal in parallel under an orchestrator. How it works, a four-agent example, and where it stops helping.
Coding agentsKeeping coding agents working after you close your laptop
What keeps background coding agents running after you close your laptop: where the agent runs, which events wake it, and what stops when a machine sleeps.
Coding agentsFrom a Linear ticket to a reviewed pull request
How a Linear coding agent turns an issue into a reviewed pull request: delegation and mentions, project routing, progress in the issue, and review before merge.
Coding agentsMulti-agent orchestration for a refactor
How to split a refactor across coding agents: map shared files, give each writer its own checkout, sequence what overlaps, and review every pull request.
Code review
What automated review checks on a pull request, and what stays with the people who merge it.
Security
Where coding agents run, what they can reach, and the evidence a security team can ask for.
SecurityGiving coding agents cloud access without stored keys
How OIDC federation gives coding agents access to AWS, Google Cloud, Azure and Vault without a stored key, and what the trust policy checks.
SecurityConnecting cloud coding agents to private APIs
How cloud coding agents reach private APIs and databases through a Tailscale tailnet: ephemeral nodes, a tag your ACLs control, and why that beats an IP allow list.
SecurityA security checklist for coding agents
A security checklist for coding agents: where they run, which credentials they hold, what they can leak, and which actions still need a person to approve.
Capy guides
How to set up common jobs in Capy, step by step, with what you should see and how to recover.
Capy guidesHow to run coding agents on your own computer and in the cloud
Run Capy threads in a folder on your own computer or on a cloud machine, follow them from any browser, and move a thread between the two without losing work.
Capy guidesHow to work through a bug backlog with coding agents
Work through a bug backlog with a Capy Captain and Linear: delegate issues, run one crew thread per bug, review each pull request, and keep conflicts small.
Capy guidesHow to fix failing CI automatically with a coding agent
Set up a Capy automation that starts when a GitHub Actions run fails on main, reads the failing job, and opens a fix pull request, with filters and recovery steps.
Capy guidesHow to run nightly dependency updates with a coding agent
Set up a Capy automation that updates dependencies every night: a schedule trigger, a prompt, a pull request that follows CI, and how to recover a failed run.
Capy guidesHow to use your own ChatGPT, Copilot or SuperGrok plan, or API keys, with Capy
Connect a ChatGPT (Codex), GitHub Copilot or SuperGrok plan, or your own API keys, to Capy: the order connections run in, what still uses credits, and limits.
Add Capy to your team
Hand Capy a task from Slack, Linear or the app, and get back a pull request ready to merge.






