Codex
31f2c2119a
2.7 KiB
Priorities
The ranked work queue for the autonomous improvement loop. The architecture-review pass (the architect) owns this file: each run it turns the roadmap plus an internal scan (gaps in the services → agents → workflows lifecycle, API coherence, drift, tech debt, test and DX friction) into a single ordered list — highest-value first — and links each item to a tracking issue. The hourly continuous-improvement pass works the top item whose issue is still open. So the architect decides what, and the increment loop builds it.
Reading / editing. An item is done when its linked issue closes (the increment
that builds it adds Closes #<issue>). Roadmap phase (Now → Next → Later) is the
primary ordering; internal findings are interleaved by value, not kept in a
separate list. The human can reorder this list — or the issues — at any time to
redirect the loop; direction always wins.
Off-limits to the loop (the architect proposes these as notes, never as queue items the loop can auto-merge): brand/positioning copy, breaking public-API changes, architectural rewrites. Those go to the human.
Work queue (ranked)
- CI-verify the first-agent 0-to-hero contract (#4272) — Current goal is developer adoption, and the README/website now describe a strong first-agent path after the examples wayfinding work (#4239/#4266), but the highest-value next increment is making that on-ramp executable as a provider-free contract. Prove scaffold → run → chat → inspect → deploy dry-run against the maintained examples so services → agents → workflows is something a new developer can walk, not just read.
- Propagate cancellation and retry signals through provider model calls (#4273) — With the AtlasCloud guarded-delegation failures closed by #4270, the next Now-phase reliability seam is real-provider failure handling: context deadlines, cancellation, retry/backoff, and rate limits must not duplicate completed tool side effects. Keep this behind focused tests/fakes and avoid public API changes.
- Add an agent debugging quickcheck to the CLI harness (#4274) — The on-ramp is only credible if the first surprising agent run is inspectable. Add provider-free harness coverage for the documented
micro inspect agent/run-inspection path and keep README/website debugging wayfinding aligned, so the inner loop does not stop at “it ran” but reaches “I can understand what happened.”
Seeded by Claude Code from the roadmap + open issues; thereafter maintained by the architecture-review pass.