# Wayari articles

- [how wayari was designed and launched with ignura](https://wayari.com/blog/how-wayari-was-launched-with-ignura/index.md) (2026-10-11): Wayari's site and launch were made with Ignura. Ignura is a launch studio for consumer apps, prosumer tools and B2B. Here is what that work covered.
- [how to use coding agents overnight without waking up to a mess](https://wayari.com/blog/how-to-use-coding-agents-overnight/index.md) (2026-10-01): A practical way to give a coding agent work for the night, with a clear finish line, safe boundaries and a pull request you can review in the morning.
- [coding agent vs ai code completion: what changes when the tool can finish the job](https://wayari.com/blog/coding-agent-vs-ai-code-completion/index.md) (2026-10-01): AI completion helps with the next line. A coding agent can investigate, edit, test and prepare a pull request. Here is when each is useful.
- [how to review ai-generated code without reading every line twice](https://wayari.com/blog/how-to-review-ai-generated-code/index.md) (2026-10-01): A practical review sequence for AI-generated pull requests: read the goal, inspect the risky seams, verify the checks and exercise the changed behaviour.
- [an AGENTS.md template for your next coding agent](https://wayari.com/blog/agents-md-practical-template/index.md) (2026-10-01): Use AGENTS.md to give coding agents the commands, boundaries and repository facts they cannot safely infer. Keep it short enough to stay true.
- [autonomous coding agents: the useful version still waits for your yes](https://wayari.com/blog/autonomous-coding-agents/index.md) (2026-10-01): Autonomous coding agents are most useful when they can plan, build, test and review a bounded job while keeping product decisions and the final merge with you.
- [at your desk, or away](https://wayari.com/blog/burst-and-sunrise/index.md) (2026-09-30): Burst for the task you want moving now. Sunrise for work you leave running. Both start in your usual coding app and end at a reviewed pull request.
- [what self improving coding should mean](https://wayari.com/blog/what-self-improving-coding-means/index.md) (2026-09-30): A correction should survive the chat where you made it. Here is what Wayari can remember for the next task, and what that does not prove.
- [how to write a brief for a coding agent](https://wayari.com/blog/how-to-write-a-brief-for-a-coding-agent/index.md) (2026-09-30): give a coding agent a problem it can reproduce, a boundary it can respect and an answer it can check. the brief should survive the first handover.
- [how to run coding agents in parallel without conflicts](https://wayari.com/blog/run-coding-agents-in-parallel-without-conflicts/index.md) (2026-09-30): separate worktrees protect files. clear ownership protects the change. decide where agents can work together before they start editing the same seam.
- [how to review an ai-generated pull request](https://wayari.com/blog/how-to-review-an-ai-generated-pull-request/index.md) (2026-09-30): read the goal beside the diff, check what actually ran and inspect the failure path. an ai-written summary is a starting point for review.
- [a green test suite never opened the page](https://wayari.com/blog/a-green-test-suite-never-opened-the-page/index.md) (2026-09-30): passing tests prove the cases they exercise. a browser check follows the person through the feature, including the part where they try again.
- [what to put in agents.md](https://wayari.com/blog/what-to-put-in-agents-md/index.md) (2026-09-30): give the next coding agent the commands and boundaries it cannot safely guess. keep repository instructions short enough to maintain when the repo changes.
- [how wayari merges](https://wayari.com/blog/why-wayari-stops-at-the-pull-request/index.md) (2026-09-22): wayari merges its own pull requests. give it a merge grant and it merges on its own the moment the checks and a clean review say so, or tell your agent to merge it.
- [the reviewer is never the author](https://wayari.com/blog/the-reviewer-is-never-the-author/index.md) (2026-09-22): once the gate goes green, an agent that did not write the change reads it and pins each finding to a line. it never hands you a verdict.
- [loop until it holds](https://wayari.com/blog/loop-until-it-holds/index.md) (2026-09-22): nothing wayari builds reaches you until it passes a gate. this is what the gate checks, and how it got fast enough to run after every stage.
- [what your repository remembers](https://wayari.com/blog/what-your-repository-remembers/index.md) (2026-09-22): every agent wayari starts is a fresh session. what carries over is a ledger of things that actually happened in your repository, each with the commit it happened at.

[HTML version](https://wayari.com/blog)