Dev open-sources minutehand, a simulator for testing proactive agents with accelerated simulated time
AshecAGE · reddit · 2026-10-21
A Reddit developer open-sourced minutehand (Apache-2.0, on PyPI), a simulator/evaluator for testing proactive agents that decide on their own when to wake up — errors that normally surface only after days.
Key ideas:
- The agent runs unchanged; a proxy routes Slack, Jira, GitHub and Google calls to fakes modeled on vendor docs and observed real-world behavior
- Simulated humans have working hours, holidays, and sometimes don't reply; the clock is simulated so long stretches run quickly
- Each run is saved as SQLite for replay, rollback to any point, and trying different prompts or models; OpenTelemetry keeps working; past runs can lock down behavior in CI
The author argues most agent projects don't need frontier models: with this test world, tuning prompts on smaller models (gpt-6-luna, open-weights) works surprisingly well because you see exactly where they slip.
More from coding & agent
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Jensen Huang details Vera Rubin agentic stack: NVL72, MGX, CMX and RTX — rohanpaul_ai · 2026-10-23
- Building an autonomous SEO agent in Hermes with three skills: programmatic pages, Keywords Everywhere, DataForSEO — EXM7777 · 2026-10-23
- Building a sports betting AI agent with onchain execution and human approval by default: architecture and lessons — Dry-Comfortable-2678 · 2026-10-23
- Capy 0.5.0 lets Claude Pro/Max subscribers reuse their plan quota, adds OIDC cloud access — garrytan · 2026-10-23
- You.com CEO Richard Socher: legacy code is human slop, let AI clean it up — RichardSocher · 2026-10-23