feat(prime): support abort_and_send_queued capability with fallback to abort

[rynfar]

Summary

Upstream Prime Agent added abort_and_send_queued to allow interrupting the current streaming turn while immediately pumping queued messages (PR #2426).

This PR adds preparatory runtime support for the negotiated capability abort_and_send_queued_v1. Pylon’s current user stop and cancellation paths still call abortAndClearQueue, so this PR does not yet expose send-queued behavior in the product:

• Adds optional abortAndSendQueued to PrimeAgentDaemonConnection in PrimeAgentDaemonBridge.ts.
• Exposes abortAndSendQueued on PrimeAgentDaemonSessionRuntime and wires abort to call connection.abortAndSendQueued() when negotiated capability abort_and_send_queued_v1 is enabled and available on the connection.
• Preserves full backward-compatible fallback to connection.abort() when the capability is not advertised or when running against stock/older daemons.
• Respects prompt lifecycle admission fencing and error handling invariants.

Testing

• Unit tests added in PrimeAgentDaemonSessionRuntime.test.ts verifying:
  • Invoking abortAndSendQueued when abort_and_send_queued_v1 is negotiated.
  • Clean fallback to abort when capability is false/absent.
  • Clean fallback to abort when capability is true but method is missing.
  • Error propagation when the connection method rejects.
• All provider test suites passing:
  • PrimeAgentDaemonBridge.test.ts (25/25 passing)
  • PrimeAgentDaemonSessionRuntime.test.ts (271/271 passing)
  • PrimeAgentDaemonAdapter.test.ts (293/293 passing)
• pnpm --filter t3 typecheck and vp lint passed with 0 errors.

Relates to fork PR pylon-code/prime-agent#81.

---

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

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated
pylon-marketing	Ignored	Preview	Sep 22, 2026 1:04pm UTC

[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	13.9 KiB	14.0 KiB	+77 B (+0.5%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.2 KiB	7.2 KiB	−4 B (−0.1%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.7 KiB	6.8 KiB	+81 B (+1.2%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	58.0 KiB	58.0 KiB	+88 B (+0.1%)	66.4 KiB	✅
Codex	Live turn messages	8	10	+2 (+25.0%)	21	✅
Claude	Total thread wire	14.0 KiB	14.0 KiB	+30 B (+0.2%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.2 KiB	7.2 KiB	+1 B (+0.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.7 KiB	6.8 KiB	+29 B (+0.4%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	58.9 KiB	58.9 KiB	+44 B (+0.1%)	66.4 KiB	✅
Claude	Live turn messages	9	10	+1 (+11.1%)	21	✅

Baseline: 84f540c · PR result: a7fb782 · 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: 115.6 KiB
• Claude decoded thread snapshot: 116.3 KiB

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