/.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. Add cross-provider agent conformance (#3901) — #3895 closed the provider deadline/retry guidance gap and #3899 closed the AP2 foundation, leaving no open codex PRs or issues in flight. The highest-value Now-phase gap is therefore battle-testing the same first-agent behavior across providers: tool calling, multi-step runs, plan/delegate, and guardrails should fail loudly by provider instead of drifting behind a single happy-path backend. Keep default CI no-secret friendly with mock coverage and key-gated live providers.
2. Checkpoint and resume durable agent runs (#3902) — adoption now has install, scaffold, run, chat, inspect/history, and 0→hero coverage, but the lived harness story still has an asymmetry: flows checkpoint and resume while long agent runs do not. This is the top Next-phase agentic-depth gap because it makes services → agents → workflows feel like one runtime under interruption and deploy/restart conditions. Keep it additive; do not break the public agent API.
3. Broaden provider streaming and keep chat/A2A streaming end to end (#3903) — after conformance and durability, streaming is the next developer-visible seam in the inner loop. Real chat and long-running A2A tasks need token streaming to stay coherent from provider → ai.Stream → micro chat → A2A message/stream, with mock/default CI coverage plus key-gated live provider checks and safe fallback for non-streaming providers.

Seeded by Claude Code from the roadmap + open issues; thereafter maintained by the architecture-review pass.