/internal/docs/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 a scheduled agent run harness contract (#3486) — with conformance, failure/cancellation hardening, checkpoint resume, memory summarization, and A2A continuity now shipped, the highest-value remaining gap is proving unattended agents as first-class operational-harness workloads. This should compose existing services, agents, flows, store-backed memory, verification, conformance, and run inspection so the scaffold → run → chat → inspect lifecycle extends to scheduled/looping work without adding a hosted scheduler, graph DSL, or breaking public APIs.

2. Export agent RunInfo as OpenTelemetry spans (#3501) — once agents can run unattended, operators need production-grade visibility that lines up with the existing tracing stack. Map run lifecycle, checkpoints/resume, tool/delegate steps, and terminal failure/cancellation metadata into OpenTelemetry rather than creating a separate observability surface.

3. Broaden provider-backed ai.Stream conformance (#3502) — A2A/chat streaming is now a visible UX seam, but the trust story depends on every provider adapter behaving consistently for streaming deltas, cancellation, and errors. Keep this as a conformance extension with mock/no-secret coverage plus provider-gated checks, so interop hardening stays CI-verifiable.

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