feat: Create Command Behavior Verification Matrix across CLI, TUI, and…

SF-Task: S04/T01

.agents/prompts/snippets/non-goals.md

@@ -1,10 +0,0 @@
-# Non-Goals — quick reference
-
-For the full set see [`.sf/NON-GOALS.md`](../../../.sf/NON-GOALS.md).
-
-- Not a code-completion product. SF is a planning + workflow control
-  plane; the agent generates code, not autocomplete suggestions.
-- Not a chatbot. Sessions are structured around milestones and modes.
-- Not a knowledge graph product. Memory and recall are operational
-  concerns, not the product itself.
-- Not a registry. SF doesn't host a public skill/plugin marketplace.

.agents/prompts/snippets/principles.md

@@ -1,13 +0,0 @@
-# Principles — quick reference
-
-For the full set see [`.sf/PRINCIPLES.md`](../../../.sf/PRINCIPLES.md).
-
-- **Purpose-first.** Every milestone, slice, task ties back to a
-  stated purpose. Refuse work that doesn't.
-- **Promote-only state.** `.sf/` artifacts are append/promote — don't
-  rewrite history of promoted artifacts; create new revisions if
-  something changed.
-- **Spec-first TDD.** Write the contract before the code; tests
-  before the implementation.
-- **Deterministic resolution.** Given the same inputs, agents must
-  produce the same canonical configuration.

.agents/prompts/snippets/style.md

@@ -1,14 +0,0 @@
-# Style — quick reference
-
-For the full guide see [`.sf/STYLE.md`](../../../.sf/STYLE.md).
-
-- DB access via `node:sqlite` `DatabaseSync`. Never `better-sqlite3`,
-  never native addons.
-- No file-based proxies for DB state. Query the DB.
-- Comments are sparing — explain WHY a non-obvious choice was made,
-  never WHAT the code does (the code says that).
-- Functions/classes get docstrings only when their purpose is
-  non-obvious from the name.
-- TypeScript-first for new code under `src/`. JavaScript-first under
-  `src/resources/extensions/sf/` (loaded directly, no compile step
-  beyond `copy-resources`).

.sf/model-performance.json

@@ -1,14 +1,14 @@
 {
   "research-slice": {
     "kimi-coding/kimi-k2.6": {
-      "successes": 4,
+      "successes": 5,
       "failures": 0,
       "timeouts": 0,
-      "totalTokens": 1590810,
-      "totalCost": 0.22167976,
-      "lastUsed": "2026-05-08T13:36:05.865Z",
+      "totalTokens": 4682735,
+      "totalCost": 0.25357486,
+      "lastUsed": "2026-05-12T20:54:43.410Z",
       "successRate": 1,
-      "total": 4
+      "total": 5
     },
     "minimax/MiniMax-M2.7": {
       "successes": 1,
@@ -51,6 +51,16 @@
       "lastUsed": "2026-05-10T07:26:04.505Z",
       "successRate": 1,
       "total": 2
+    },
+    "kimi-coding/kimi-k2.6": {
+      "successes": 1,
+      "failures": 0,
+      "timeouts": 0,
+      "totalTokens": 476746,
+      "totalCost": 0,
+      "lastUsed": "2026-05-12T20:56:35.833Z",
+      "successRate": 1,
+      "total": 1
     }
   },
   "discuss-milestone": {
@@ -97,6 +107,16 @@
       "lastUsed": "2026-05-10T15:16:08.120Z",
       "successRate": 1,
       "total": 1
+    },
+    "kimi-coding/kimi-k2.6": {
+      "successes": 1,
+      "failures": 0,
+      "timeouts": 0,
+      "totalTokens": 1821480,
+      "totalCost": 0,
+      "lastUsed": "2026-05-12T20:57:45.179Z",
+      "successRate": 1,
+      "total": 1
     }
   },
   "complete-slice": {

.sf/safety/evidence-M001-6377a4-S04-T01.json

@@ -0,0 +1,16 @@
+[
+  {
+    "kind": "write",
+    "toolCallId": "write_1778619443353_32",
+    "path": ".sf/milestones/M001-6377a4/slices/S04/VERIFICATION_MATRIX.md",
+    "timestamp": 1778619443535
+  },
+  {
+    "kind": "bash",
+    "toolCallId": "bash_1778619447339_33",
+    "command": "test -f .sf/milestones/M001-6377a4/slices/S04/VERIFICATION_MATRIX.md && grep -q \"status\" .sf/milestones/M001-6377a4/slices/S04/VERIFICATION_MATRIX.md && echo \"Matrix exists and contains status command info.\"",
+    "exitCode": 0,
+    "outputSnippet": "Matrix exists and contains status command info.\n",
+    "timestamp": 1778619447544
+  }
+]

.sf/traces/guard:76c7c307-91b4-426e-8fad-4ff951d5a52e.jsonl

@@ -0,0 +1 @@
+{"ts":"2026-05-12T20:56:39.435Z","type":"gate_run","traceId":"guard:76c7c307-91b4-426e-8fad-4ff951d5a52e","turnId":"iter-4","gateId":"plan-gate","gateType":"policy","unitType":"execute-task","unitId":"M001-6377a4/S04/T01","milestoneId":"M001-6377a4","sliceId":"S04","outcome":"pass","failureClass":"none","rationale":"Plan files verified","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:56:39.435Z","durationMs":0}

.sf/traces/latest

@@ -1 +1 @@
-pre-dispatch:952f3f1c-1861-4637-a706-54141371ebc7.jsonl
+guard:76c7c307-91b4-426e-8fad-4ff951d5a52e.jsonl

.sf/traces/pre-dispatch:2e5b5158-453c-4b09-8dd2-26fa87c8ded5.jsonl

@@ -0,0 +1,2 @@
+{"ts":"2026-05-12T20:52:04.171Z","type":"gate_run","traceId":"pre-dispatch:2e5b5158-453c-4b09-8dd2-26fa87c8ded5","turnId":"iter-1","gateId":"resource-version-guard","gateType":"policy","unitType":"pre-dispatch","unitId":"iter-1","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"resource version guard passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:52:04.167Z","durationMs":0}
+{"ts":"2026-05-12T20:52:04.988Z","type":"gate_run","traceId":"pre-dispatch:2e5b5158-453c-4b09-8dd2-26fa87c8ded5","turnId":"iter-1","gateId":"pre-dispatch-health-gate","gateType":"execution","unitType":"pre-dispatch","unitId":"iter-1","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"pre-dispatch health gate passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:52:04.987Z","durationMs":0}

.sf/traces/pre-dispatch:76c7c307-91b4-426e-8fad-4ff951d5a52e.jsonl

@@ -0,0 +1,3 @@
+{"ts":"2026-05-12T20:56:38.487Z","type":"gate_run","traceId":"pre-dispatch:76c7c307-91b4-426e-8fad-4ff951d5a52e","turnId":"iter-4","gateId":"resource-version-guard","gateType":"policy","unitType":"pre-dispatch","unitId":"iter-4","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"resource version guard passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:56:38.485Z","durationMs":0}
+{"ts":"2026-05-12T20:56:39.307Z","type":"gate_run","traceId":"pre-dispatch:76c7c307-91b4-426e-8fad-4ff951d5a52e","turnId":"iter-4","gateId":"pre-dispatch-health-gate","gateType":"execution","unitType":"pre-dispatch","unitId":"iter-4","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"pre-dispatch health gate passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:56:39.307Z","durationMs":0}
+{"ts":"2026-05-12T20:56:39.313Z","type":"gate_run","traceId":"pre-dispatch:76c7c307-91b4-426e-8fad-4ff951d5a52e","turnId":"iter-4","gateId":"planning-flow-gate","gateType":"policy","unitType":"pre-dispatch","unitId":"iter-4","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"planning flow compile gate passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:56:39.312Z","durationMs":0}

.sf/traces/pre-dispatch:7b06b885-0539-40d2-81c1-066c2b2f20c5.jsonl

@@ -0,0 +1,2 @@
+{"ts":"2026-05-12T20:56:36.969Z","type":"gate_run","traceId":"pre-dispatch:7b06b885-0539-40d2-81c1-066c2b2f20c5","turnId":"iter-3","gateId":"resource-version-guard","gateType":"policy","unitType":"pre-dispatch","unitId":"iter-3","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"resource version guard passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:56:36.968Z","durationMs":0}
+{"ts":"2026-05-12T20:56:37.702Z","type":"gate_run","traceId":"pre-dispatch:7b06b885-0539-40d2-81c1-066c2b2f20c5","turnId":"iter-3","gateId":"pre-dispatch-health-gate","gateType":"execution","unitType":"pre-dispatch","unitId":"iter-3","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"pre-dispatch health gate passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:56:37.700Z","durationMs":0}

.sf/traces/pre-dispatch:9ed021af-f7d3-4d67-bfae-5ecbfcbd08b6.jsonl

@@ -0,0 +1,2 @@
+{"ts":"2026-05-12T20:54:44.467Z","type":"gate_run","traceId":"pre-dispatch:9ed021af-f7d3-4d67-bfae-5ecbfcbd08b6","turnId":"iter-2","gateId":"resource-version-guard","gateType":"policy","unitType":"pre-dispatch","unitId":"iter-2","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"resource version guard passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:54:44.466Z","durationMs":0}
+{"ts":"2026-05-12T20:54:45.217Z","type":"gate_run","traceId":"pre-dispatch:9ed021af-f7d3-4d67-bfae-5ecbfcbd08b6","turnId":"iter-2","gateId":"pre-dispatch-health-gate","gateType":"execution","unitType":"pre-dispatch","unitId":"iter-2","milestoneId":"M001-6377a4","outcome":"pass","failureClass":"none","rationale":"pre-dispatch health gate passed","findings":"","attempt":1,"maxAttempts":1,"retryable":false,"evaluatedAt":"2026-05-12T20:54:45.217Z","durationMs":0}