/.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. CI-verify first-agent CLI wayfinding commands (#4336) — Developer adoption is the current goal, and the README/docs now steer new users through installed CLI affordances (micro agent demo, micro examples, micro zero-to-hero) before deeper tutorials. Those commands should be treated as a no-secret on-ramp contract so the binary keeps pointing developers to the maintained first-agent, debugging, examples, and 0→hero path instead of drifting silently.
2. Add durable checkpoint/resume for agent runs (#4341) — With the plan/delegate recovery issue closed, the next biggest lifecycle seam is that flows can checkpoint and resume but long agent runs still need a durable recovery contract. Solving this keeps the services → agents → workflows story cohesive without replaying completed tool side effects, and it should stay scoped to a non-breaking, CI-verifiable harness increment.
3. Trace agent RunInfo in OpenTelemetry spans (#4315) — Once the CLI on-ramp is protected and agent runs have a durable recovery path like flows, the highest Next-phase operability gap is connecting existing run metadata to traces so real agent runs can be debugged across steps, tool calls, delegation, failures, services, and flows without inventing a new surface.

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