[gHashTag]

M4 of T27F Live (epic #10266), issue #11778. ONE new spec: specs/t27f/live_mode.t27.

• FR-001 structural: live_frame goes through the same work_graph apply_event -> display_list flatten path as the M3 replay; the fixture proves live-frame-by-frame equals batch replay byte- and sha-equal per frame and by final digest. No second renderer.
• Coalescing without loss: a burst of 10 events across two windows paints once; the graph stays byte-identical to one-by-one replay; all 10 action nodes present.
• FR-002: MeasureRecord (seq/emitted_at/painted_at/latency) full rows; p50/p99 constants 50/150 ms stated; a run past the p99 budget is REFUSED (error, never silent).
• Order discipline per live_stream: out-of-order / repeat / gap seq refused.
• Honest scope: the headless-browser lab measurement is NOT CLAIMED; LAB_RUN_DONE=false asserted (the T27 OS lab contract is t27os specs/os/lab.t27, infra: the lab leaves Railway -- run drive-in and receipts re-homed to T27 OS (github-actions executor, t27os runs/ publisher) #11781).

8/8 tests pass, 0 vacuous; sealed (all hashes MATCH, re-sealed with the current-source t27c); dupe ledger blessed in the same commit (the deliberate node_of local copy). Reuses apply_event/flatten_into/stream_empty -- nothing restated.

phi^2 + 1/phi^2 = 3 | TRINITY

[github-actions]

PR Dashboard

Generated at: 2026-10-23 11:23:18 UTC

• spec(t27f): live_mode -- live SSE event to painted frame, M4 of T27F Live (Closes #11778) #11787 spec(t27f): live_mode -- live SSE event to painted frame, M4 of T27F Live (Closes #11778) | Created: 2026-10-23 | Updated: 2026-10-23 | 28 failing / 28 total
• fix(gates): master reds after the port-train wave -- types ratchet, seals (Refs #11038) #11786 fix(gates): master reds after the port-train wave -- types ratchet, seals (Refs #11038) | Created: 2026-10-23 | Updated: 2026-10-23 | 8 failing / 28 total
• t27f: hud + a11y_mirror -- agent-state HUD and its text mirror (Closes #10262) #11782 t27f: hud + a11y_mirror -- agent-state HUD and its text mirror (Closes #10262) | Created: 2026-10-23 | Updated: 2026-10-23 | 0 failing / 2 total
• ledger: face.t27 is t27b pass, as the lab measured #11699 ledger: face.t27 is t27b pass, as the lab measured | Created: 2026-10-22 | Updated: 2026-10-23 | 6 failing / 27 total
• dupe ledger: bless bytes_eq group (master Duplicate Body Ratchet green) #11426 dupe ledger: bless bytes_eq group (master Duplicate Body Ratchet green) | Created: 2026-10-20 | Updated: 2026-10-23 | 4 failing / 22 total
• tri queen tick: the local Queen -- P0/P1 picked, claimed on origin, run as lanes on any Mac (Closes #11415) #11420 tri queen tick: the local Queen -- P0/P1 picked, claimed on origin, run as lanes on any Mac (Closes #11415) | Created: 2026-10-20 | Updated: 2026-10-23 | 14 failing / 36 total
• gen-rust: refuse a []T struct field filled from borrowed storage (D2 of #11182, needs the ledger decision of #11218) #11231 gen-rust: refuse a []T struct field filled from borrowed storage (D2 of #11182, needs the ledger decision of #11218) | Created: 2026-10-20 | Updated: 2026-10-23 | 11 failing / 38 total
• check: refuse a module that declares a top-level name twice (Closes #10421) #10442 check: refuse a module that declares a top-level name twice (Closes #10421) | Created: 2026-10-18 | Updated: 2026-10-23 | 10 failing / 38 total
• spec(port): share tree helpers across eight gen_w ports (Closes #9633) #9636 spec(port): share tree helpers across eight gen_w ports (Closes #9633) | Created: 2026-10-15 | Updated: 2026-10-23 | 6 failing / 23 total
• gen-zig: a one-element array constant ([1]u32 = [48]) makes test-report BLOCK trinity specs/reproduce/contracts.t27 #6958 gen-zig: a one-element array constant ([1]u32 = [48]) makes test-report BLOCK trinity specs/reproduce/contracts.t27 | Created: 2026-10-06 | Updated: 2026-10-23 | 15 failing / 38 total
• A test or bench body that assigns a module-level var gets a shadowing local, so the spec does not compile #6833 A test or bench body that assigns a module-level var gets a shadowing local, so the spec does not compile | Created: 2026-10-06 | Updated: 2026-10-23 | 12 failing / 38 total
• bitwalk: --help and a missing file panic instead of printing usage or an error #6551 bitwalk: --help and a missing file panic instead of printing usage or an error | Created: 2026-10-05 | Updated: 2026-10-23 | 9 failing / 29 total
• Port gHashTag/BrowserOS:trios/agent-server/apps/server/src/api/services/queen-brief-shape.ts (TypeScript, 2 functions) t #4911 Port gHashTag/BrowserOS:trios/agent-server/apps/server/src/api/services/queen-brief-shape.ts (TypeScript, 2 functions) t | Created: 2026-09-27 | Updated: 2026-10-23 | 10 failing / 29 total
• Implement the 8 empty function bodies in specs/vsa/packed_vsa.t27 #4729 Implement the 8 empty function bodies in specs/vsa/packed_vsa.t27 | Created: 2026-09-24 | Updated: 2026-10-23 | 8 failing / 29 total
• Wave 684: occlusion four deep, and the two files no list ever reached #4660 Wave 684: occlusion four deep, and the two files no list ever reached | Created: 2026-09-23 | Updated: 2026-10-23 | 17 failing / 69 total
• Capture given/when/then and render them: 6540 clauses, and 169 specs share one unfilled placeholder #4642 Capture given/when/then and render them: 6540 clauses, and 169 specs share one unfilled placeholder | Created: 2026-09-23 | Updated: 2026-10-23 | 18 failing / 81 total
• Implement part 2 of 2 of the empty bodies in specs/git/diff.t27 (7 functions) #4634 Implement part 2 of 2 of the empty bodies in specs/git/diff.t27 (7 functions) | Created: 2026-09-23 | Updated: 2026-10-23 | 18 failing / 81 total
• Port part 1 of 2 of tools/check_withdrawn_live.py to specs/port/tools/check_withdrawn_live.t27 (8 functions) #4581 Port part 1 of 2 of tools/check_withdrawn_live.py to specs/port/tools/check_withdrawn_live.t27 (8 functions) | Created: 2026-09-22 | Updated: 2026-10-23 | 17 failing / 69 total
• Implement part 2 of 2 of the empty bodies in specs/vm/jit_semantics.t27 (4 functions) #4563 Implement part 2 of 2 of the empty bodies in specs/vm/jit_semantics.t27 (4 functions) | Created: 2026-09-21 | Updated: 2026-10-23 | 17 failing / 69 total

Summary

Status	Count
Total Open PRs	19
PRs with Failing Checks	17
PRs with All Checks Green	2
READY	1
FAILING	17
PENDING	0
NO CHECKS YET	0

These columns do not partition: 1 + 17 + 0 + 0 = 18, and there are 19 open PRs. A PR is being counted twice or not at all.

Seal Status

• ⚠️ STALE -- sha256(compiler.rs)=5aec047d42ec != manifest seal=87e5cbd3ad94.
  The committed NMSE numbers were certified against an older compiler.rs.
  Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

[tri-claw]

Summary by t27-bees

Checked by the t27 compiler

_{t27c parse, parse-complete and typecheck on each changed .t27 file at a890ce6. Nothing was generated, built or run. A pass means the file is well formed, not that it is right.}

• specs/t27f/live_mode.t27: parses and typechecks.

Read by the model

This spec adds M4 live mode: a single SSE event flows through work_graph::apply_event → display_list::flatten_into (the same path as M3 replay), with coalescing (one paint per frame window, no graph drops), per-event latency measurement, and strict order/budget gates. The seal shows 8/8 tests passing, 0 vacuous.

Worth a look

• specs/t27f/live_mode.t27:123 — Local copy of node_of duplicates work_graph.t27's private function because queen::actors.t27 exports a different node_of. If work_graph changes its lookup (scan order, return value, node kinds), this copy silently diverges; the FR-001 test (digests_equal(graph_digest(&ga), graph_digest(&gb))) would catch it, but the duplication is a maintenance hazard.

• specs/t27f/live_mode.t27:197 — fixture_scene asserts agents <= SCENE_AGENT_CAP (20). This is a pub function; if called with a graph holding >20 agents (e.g., from a future test or integration), it crashes rather than returning an error code. The comment says "never truncated" but the assert makes it a hard crash.

• specs/t27f/live_mode.t27:221-222 — paint_frame resets st.n = 0; st.ops = 0 before calling flatten_into. If flatten_into returns an error, the caller's stream is already cleared (bytes zeroed). The comment claims "a refusal before this point leaves the caller's stream byte-identical", but the reset happens at this point.

• specs/t27f/live_mode.t27:285 — percentile takes a slice a: []u32 and sorts it in place (via sort_u32), mutating the caller's data. run_within_budget copies to a local buffer first, but the function signature doesn't signal mutation.

• specs/t27f/live_mode.t27:36 — MAX_NODES = 32 is a faithful copy of scene.t27's cap, pinned here because spec_graph.t27 declares a different MAX_NODES (1024) in the same import closure. If scene.t27 changes its cap, this constant and the test at line 106 (apply_op refuses 33rd node) must be updated manually; the duplicate-bodies baseline now tracks node_of but not this constant.

---

_{Read: 3 files, +877 -0, the whole diff. Head a890ce6. Model: openai-compatible/nvidia/nemotron-3-ultra-550b-a55b. Commands: @t27-bees review | summary | help | pause | resume. Free for every repository.}