fix(server): keep OpenCode approvals, progress, and Stop responsive

[rynfar]

OpenCode can appear stuck when its output pipes fill, an approval has no usable controls, or Stop finishes before delegated sessions stop. Lost reply events and disconnects can also leave approvals and running indicators stale.

Finish the previously started adoption of upstream #9005, #9282, and #9653. Keep output pipes drained, make permission replies bounded and retryable, keep automatic replies off the event pump, reconcile missing requests after reconnect, and wait for child cleanup on Stop. Native task progress updates only an active parent turn; duplicate plans are suppressed and cancelled tasks are omitted, completing the intent of #178.

Preserve Pylon's session-incarnation and turn-admission guards, compaction events, approval contracts, branding, and other providers. Explain OpenCode's workspace-wide permission scope through the existing detail field so web, desktop, and mobile all display it. Shared projection fixes restore failed approvals and finalize aborted text; local, external-server, remote, and tunnel connections use the same behavior.

Validation: all 400 focused tests across nine adapter, runtime, orchestration, web, mobile, and shared presentation files pass (including a real subprocess writing 2 MiB to each output pipe). Server, web, mobile, and client-runtime typechecks pass; targeted lint, formatting, and diff checks pass. No live provider session or production database was changed. Browser verification has not been run.

Implemented with GPT-6 Astra through the Codex harness, completing the earlier port and incorporating the task-progress work authored with Claude Opus 5 in Pylon.

---

^{Need help on this PR? Tag @codesmith-bot with what you need. Autofix is disabled.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.6 KiB	13.6 KiB	−2 B (−0.0%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.1 KiB	7.1 KiB	+2 B (+0.0%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.6 KiB	6.6 KiB	−4 B (−0.1%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	56.5 KiB	56.5 KiB	0 B (0.0%)	66.4 KiB	✅
Codex	Live turn messages	8	8	0 (0.0%)	21	✅
Claude	Total thread wire	13.6 KiB	13.6 KiB	0 B (0.0%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.1 KiB	7.1 KiB	−6 B (−0.1%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.6 KiB	6.6 KiB	+6 B (+0.1%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	57.3 KiB	57.3 KiB	0 B (0.0%)	66.4 KiB	✅
Claude	Live turn messages	8	8	0 (0.0%)	21	✅

Baseline: 9afb259 · PR result: 7d4c455 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 111.2 KiB
• Claude decoded thread snapshot: 111.9 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.