Qwen Code PR adds one-command local daemon and managed agent with embedded H2
doudouOUC · ghdev · 2026-10-21
A new experimental PR (#14928, 7.6k lines across 75 files) in QwenLM/qwen-code adds qwen serve --managed --open, starting both the regular daemon and a Managed Agent locally with a single foreground command.
Key points:
- Serves an authenticated unified Web Shell and APIs over a loopback origin; Java management routes stay private
- Uses persistent embedded H2 with a private Hosted Harness — no MySQL, Docker, Maven, or systemd required; the CLI downloads and verifies its exact JAR from official OSS on first launch, then runs offline
- Local conversation models connect to the existing Java catalog and new-task presets, with credentials kept in Harness config
- Signed Maven Central publication ensures identical JAR/checksum bytes across Central, GitHub Releases, and OSS
More from coding & agent
- Claude cut a 2.6-hour video export to 20 minutes — but only after being told to rebuild the server — byhewar · 2026-10-23
- 3 flight-booking rules my AI agent checks that I'd never do by hand — Efistoffeles · 2026-10-23
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 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
- How do you audit AI agent decisions in regulated workflows? A field survey — Ok_Way_1993 · 2026-10-23