Stacks/Coding agent
The agent teams that build AWS Emulator
AWS Emulator is written entirely by agents in three teams: one writes fidelity tests against the real AWS cloud, one finds where the emulator diverges and files issues, and one fixes them. Each fix passes a gate of compilation plus reviewer agents for standards, security, and fidelity before merging. It started as a weekend Kiro experiment.
- Use case
- Autonomous agents that write, test, and fix every line of AWS Emulator, a downloadable local emulator of AWS services, with no human-written or human-reviewed code
- Category
- Coding agent
- Team size
- not stated
- Orchestrator / framework
- Three teams of agents: one writes fidelity tests from public AWS API models and verifies each against real AWS; one replays the tests against the emulator and files grouped issues for failures; one fixes single issues. Fixes pass a gate (compilation plus agents that review coding standards, security, and fidelity regressions) or go back for rework. Hundreds of agents work concurrently and fix thousands of issues a day.
- Memory
- not stated
- Evals
- Tens of thousands of fidelity tests checked against the real AWS cloud are the main guardrail, and the merge gate blocks any change that regresses fidelity.
- Observability
- not stated
- Cost (stated in writeup)
- Not stated
- Cost per output (stated)
- Not stated in the writeup
- WhatWeRun estimate
WhatWeRun estimate
Not enough info to estimate
The post doesn't name the models or give tokens per fix; 'thousands of issues a day' isn't precise enough to price.
Reviewed Nov 13, 2026
- Observability tools
- not stated
- Other tools
- AWS CodeCommit (primary repository)AWS Developer Tools release pipelineGitHub (public code release)
- Source
- https://www.aboutamazon.com/news/aws/how-autonomous-ai-agents-built-aws-emulator
Matthew Goldberg · Published Nov 11, 2026 · Last verified Nov 13, 2026
Related
More like this
corezoid.comCorezoid's ticket-routing coding agent
A coding agent triggered by Jira labels that fixes tickets and opens merge requests, with a digital twin of the engineering org deciding which repository, process, or service each ticket belongs to
Stated: ~$2.50 per code change+1
blog.duolingo.comDuolingo AI STR Fix bug agent
Investigating and fixing Shake-to-Report bug tickets: gathering context (relevant code, tracking events, experiment states, recent PRs, repro steps) and proposing a fix for engineers to verify
lovable.devLovable's app-building agent
Chat-driven creation, hosting, and iteration of web apps for people who can't code; users have created more than 70 million projects