# How Ryan Carson actually uses coding agents

Founder of Untangle

**If you take one thing:** `!land playbook`, refreshes skills on every PR. his Devin playbook for every merge starts a child session that updates skills and docs, so they never go stale.

- **Agent:** Devin (web) ([source, 2026-08-12](https://x.com/ryancarson/status/2087538698543907048))
- **Agent:** Codex ([source, 2026-05-22](https://x.com/ryancarson/status/2057798991086014655))
- **Model:** Claude Opus 5.5 ([source, 2026-09-26](https://x.com/ryancarson/status/2103926098081763798))
- **Model:** SWE 1.7 ([source, 2026-07-14](https://x.com/ryancarson/status/2076975558756339847))
- **Review:** Agents review and land; he watches a test video first ([source, 2026-06-01](https://x.com/ryancarson/status/2061393372737085816))
- **Harness:** Devin cloud sessions sorted into P0/P1/Bugs folders, Linear, Opus 5.5 as manager, zero local dev ([source, 2026-09-26](https://x.com/ryancarson/status/2103926098081763798))
- **Hooks:** pre-commit design-system linter ([source, 2026-03-03](https://x.com/ryancarson/status/2028901454996447296))
- **MCP servers:** next-devtools ([source, 2026-02-17](https://x.com/ryancarson/status/2023886869231239485))

## In their words

> I would say 95% of my engineering is in the cloud now with @DevinAI, and I assume it will eventually be 100%.

> 3. frontend eng local = codex

> Engineering manager Opus 5.5 + @DevinAI + @linear

> Did almost all my engineering on SWE 1.7 yesterday.

> Agents write/review/land 100% of our code now.


Archetype (our label): Orchestrator. Newest source: 2026-09-26.
Card: https://www.prehistoriq.com/stacks/snarktank/ · Data: https://github.com/prehistoriq/stacks/blob/main/data/stacks/snarktank.json