/.github/loop/PRIORITIES.md

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)

1. Stabilize AtlasCloud plan-delegate unfinished plan steps (#4013) — #4015 closed the guarded-delegate AtlasCloud seam (#4007), and there are no open codex PRs currently carrying builder work. The remaining open Now-phase conformance gap is the live AtlasCloud/minimax plan-delegate harness: it can stop after the first side effect with unfinished plan steps instead of completing the services → agents → workflows path (three tasks plus one delegated notification). Keep this first because the loop's evaluator depends on provider conformance staying trustworthy, and the fix must preserve the checkpoint guard rather than weakening unfinished-plan detection.
2. Trace agent runs as OpenTelemetry spans (#3908) — the README, website, and recent blog story align around the operational harness: services become tools, agents are services, flows handle deterministic paths, and developers should be able to scaffold → run → chat → inspect → deploy. With first-agent wayfinding, no-secret harnesses, and streaming now represented, the next highest-value developer-visible seam is observability: RunInfo steps, tool calls, delegation, status, duration, and failures should correlate to spans while defaulting to no-op when tracing is not configured. Seeded by Claude Code from the roadmap + open issues; thereafter maintained by the architecture-review pass.