All ideasMore in Work
Work

Make claude.ai three times faster in a two-week agent sprint

Reporting Anthropic's September 23 engineering post, this analysis describes the AI running a two-week performance sprint from a Slack channel: 150-plus threads, benchmarks built for each bottleneck, and more than 3,000 merged changes with no customer-facing incident. Time to a typeable page at the 75th percentile fell from 3.1 seconds to 0.55, with a human approving every pull request.

How it works

  1. Put the AI in a Slack channel with standing instructions
  2. Analyze usage data via the Datadog MCP server
  3. Pick the four highest-impact user journeys
  4. Build a lab benchmark per bottleneck
  5. Open risk-sized pull requests behind feature flags
  6. Watch deploys and field data
  7. Ratchet each benchmark down in CI

Why it’s worth knowing

The agent ran its own maker's product-speed program: 3,000 merged changes driven from a chat channel.

Where it’s from

Summarized from a public post on fourweekmba.com. See the original ↗

Upgraded prompt

A ready-to-use prompt for this idea, built from research on what makes AI agents do the job right: who the AI should be, the exact steps, and a check to make sure it was done the way you wanted.

1Who the AI should be
2The prompt, with the steps
3The check when it’s done
Upgraded prompts are coming soon.

Try next

WorkRead every job advert, tailor my CV and apply with my approvalWorkFind me 48 school-district customers and draft each pitch — one replied, booked and paidWorkMatch 904 insurance documents overnight at my dental clinicWorkBuild a custom order-routing app that sends my worst orders to Amazon MCF

Every idea is a real post from someone who tried it, summarized in our words, with a link to the original. App names are swapped for plain words.