@ngockhoale/ukit 2.7.1 → 2.7.3

package/scripts/perf/perf-findings.json

@@ -0,0 +1,420 @@
+{
+  "generatedAt": "2026-09-20T07:42:56.112Z",
+  "schema": "ukit-perf-findings/1",
+  "sources": {
+    "telemetry": {
+      "dir": "/Volumes/KHOA_EXTENAL/WORKING_PROJECT/WORKING/UKit/.ukit/storage/cache/hook-latency",
+      "rows": 34629,
+      "files": 55
+    },
+    "settings": "/Volumes/KHOA_EXTENAL/WORKING_PROJECT/WORKING/UKit/.worktrees/task-233/templates/.claude/settings.json",
+    "hooksDir": "/Volumes/KHOA_EXTENAL/WORKING_PROJECT/WORKING/UKit/.worktrees/task-233/templates/.claude/hooks",
+    "benchmarkIterations": 10,
+    "gitLogWindowDays": 14
+  },
+  "findings": [
+    {
+      "id": "chain-pretooluse-read-grep-glob",
+      "area": "hook chain — PreToolUse Read|Grep|Glob — 1 hook",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=974, p50 57ms, 2 hook rows/call; paths: 465 direct / 509 chain-runner). Hooks: sensitive-data-guard.sh (8s, 55ms)",
+      "p50ms": 57,
+      "p95ms": 76,
+      "totalMs": 53616,
+      "invocations": 974,
+      "rootCause": "3 process spawns per call (1 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-pretooluse-edit-write",
+      "area": "hook chain — PreToolUse Edit|Write — 6 hooks",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=533, p50 316ms, 6 hook rows/call; paths: 458 direct / 75 chain-runner). Hooks: protect-files.sh (8s, 38ms), stale-spec-guard.sh (11s, 36ms), pre-edit-backup.sh (8s, 36ms), skill-router.sh (20s, 90ms), handoff-model-guard.sh (8s, 37ms), context-hardcap-gate.sh (8s, 54ms)",
+      "p50ms": 316,
+      "p95ms": 433,
+      "totalMs": 179919,
+      "invocations": 533,
+      "rootCause": "18 process spawns per call (6 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-pretooluse-bash",
+      "area": "hook chain — PreToolUse Bash — 6 hooks",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=2973, p50 349ms, 6 hook rows/call; paths: 2711 direct / 262 chain-runner). Hooks: auto-allow-bash.sh (15s, 40ms), block-dangerous.sh (8s, 44ms), sensitive-data-guard.sh (8s, 55ms), handoff-model-guard.sh (8s, 37ms), context-hardcap-gate.sh (8s, 54ms), verification-guard.sh (8s, 56ms)",
+      "p50ms": 349,
+      "p95ms": 505,
+      "totalMs": 1070247,
+      "invocations": 2973,
+      "rootCause": "18 process spawns per call (6 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-posttooluse-read-grep-glob",
+      "area": "hook chain — PostToolUse Read|Grep|Glob — 1 hook",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=963, p50 64ms, 2 hook rows/call; paths: 454 direct / 509 chain-runner). Hooks: record-execution.sh (8s, 38ms)",
+      "p50ms": 64,
+      "p95ms": 149,
+      "totalMs": 69147,
+      "invocations": 963,
+      "rootCause": "3 process spawns per call (1 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-posttooluse-edit-write",
+      "area": "hook chain — PostToolUse Edit|Write — 3 hooks",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=523, p50 125ms, 3 hook rows/call; paths: 451 direct / 72 chain-runner). Hooks: post-edit-verify.sh (8s, 36ms), record-execution.sh (8s, 38ms), task-watchdog.sh (8s, 54ms)",
+      "p50ms": 125,
+      "p95ms": 221,
+      "totalMs": 72670,
+      "invocations": 523,
+      "rootCause": "9 process spawns per call (3 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-posttooluse-bash",
+      "area": "hook chain — PostToolUse Bash — 2 hooks",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=2852, p50 90ms, 2 hook rows/call; paths: 2591 direct / 261 chain-runner). Hooks: compress-output.sh (12s, 61ms), record-execution.sh (8s, 38ms)",
+      "p50ms": 90,
+      "p95ms": 145,
+      "totalMs": 279069,
+      "invocations": 2852,
+      "rootCause": "6 process spawns per call (2 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-userpromptsubmit-all",
+      "area": "hook chain — UserPromptSubmit (all tools) — 4 hooks",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=193, p50 319ms, 4 hook rows/call; paths: 157 direct / 36 chain-runner). Hooks: sensitive-data-guard.sh (8s, 55ms), skill-router.sh (20s, 90ms), vision-router.sh (12s, 54ms), context-window-guard.sh (10s, 54ms)",
+      "p50ms": 319,
+      "p95ms": 585,
+      "totalMs": 65511,
+      "invocations": 193,
+      "rootCause": "12 process spawns per call (4 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-stop-all",
+      "area": "hook chain — Stop (all tools) — 1 hook",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=113, p50 81ms, 1 hook rows/call; paths: 113 direct / 0 chain-runner). Hooks: completion-gate.sh (8s, 36ms)",
+      "p50ms": 81,
+      "p95ms": 108,
+      "totalMs": 9542,
+      "invocations": 113,
+      "rootCause": "3 process spawns per call (1 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-precompact-all",
+      "area": "hook chain — PreCompact (all tools) — 1 hook",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=2, p50 49ms, 1 hook rows/call; paths: 0 direct / 2 chain-runner). Hooks: reinject-context.sh (8s, 33ms)",
+      "p50ms": 49,
+      "p95ms": 49,
+      "totalMs": 84,
+      "invocations": 2,
+      "rootCause": "3 process spawns per call (1 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 is under the 50ms hotspot bar; consolidation is still the mechanism that removes the spawns.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-sessionstart-all",
+      "area": "hook chain — SessionStart (all tools) — 4 hooks",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured per-call hook cost for the group's matching tools (n=58, p50 96ms, 2 hook rows/call; paths: 37 direct / 21 chain-runner). Hooks: project-important.sh (8s, 35ms), auto-prune-bash.sh (12s, 1ms), reset-compact-pressure.sh (12s, 2ms), handoff-resume.sh (8s, 54ms)",
+      "p50ms": 96,
+      "p95ms": 191,
+      "totalMs": 7170,
+      "invocations": 58,
+      "rootCause": "12 process spawns per call (4 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Measured p50 exceeds the 50ms hotspot bar, so the spawn multiplier is the dominant term.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "chain-sessionend-all",
+      "area": "hook chain — SessionEnd (all tools) — 1 hook",
+      "evidence": "settings.json",
+      "evidenceDetail": "settings.json group only — no telemetry rows are attributable to this group, so no per-call cost is reported. Hooks: session-episode.sh (8s, 23ms)",
+      "p50ms": null,
+      "p95ms": null,
+      "totalMs": null,
+      "invocations": 0,
+      "rootCause": "3 process spawns per call (1 settings rows × 3: wrapper bash + node runtime + advisory telemetry node child), against a measured floor of bash 3.8ms + node 16.9ms + telemetry finish 19.5ms. Not measured on this tree.",
+      "fixTask": null,
+      "status": "deferred"
+    },
+    {
+      "id": "hook-hook-chain-runner",
+      "area": "single hook — hook-chain-runner",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 not benchmarked (script not resolvable from this root)",
+      "p50ms": 106,
+      "p95ms": 409,
+      "totalMs": 283464,
+      "invocations": 1753,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-skill-router-sh",
+      "area": "single hook — skill-router.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 89.6ms",
+      "p50ms": 55,
+      "p95ms": 130,
+      "totalMs": 50744,
+      "invocations": 770,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-vision-router-sh",
+      "area": "single hook — vision-router.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 54.4ms",
+      "p50ms": 82,
+      "p95ms": 232,
+      "totalMs": 22275,
+      "invocations": 215,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-context-window-guard-sh",
+      "area": "single hook — context-window-guard.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 54.1ms",
+      "p50ms": 55,
+      "p95ms": 90,
+      "totalMs": 12101,
+      "invocations": 215,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-auto-allow-bash-sh",
+      "area": "single hook — auto-allow-bash.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 39.7ms",
+      "p50ms": 64,
+      "p95ms": 96,
+      "totalMs": 196059,
+      "invocations": 2994,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-block-dangerous-sh",
+      "area": "single hook — block-dangerous.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 44.3ms",
+      "p50ms": 89,
+      "p95ms": 190,
+      "totalMs": 392109,
+      "invocations": 4263,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-context-hardcap-gate-sh",
+      "area": "single hook — context-hardcap-gate.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 53.9ms",
+      "p50ms": 67,
+      "p95ms": 93,
+      "totalMs": 241749,
+      "invocations": 3571,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-stale-spec-guard-sh",
+      "area": "single hook — stale-spec-guard.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 36.0ms",
+      "p50ms": 53,
+      "p95ms": 74,
+      "totalMs": 28844,
+      "invocations": 544,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-pre-edit-backup-sh",
+      "area": "single hook — pre-edit-backup.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 36.0ms",
+      "p50ms": 55,
+      "p95ms": 74,
+      "totalMs": 29992,
+      "invocations": 535,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-completion-gate-sh",
+      "area": "single hook — completion-gate.sh",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; standalone re-run p50 35.8ms",
+      "p50ms": 81,
+      "p95ms": 108,
+      "totalMs": 9659,
+      "invocations": 116,
+      "rootCause": "per-invocation cost is dominated by the 3-spawn hook row (wrapper bash 3.8ms + node 16.9ms + telemetry finish 19.5ms), plus this script's own work",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "hook-project-important-tail",
+      "area": "project-important.sh — PROJECT_IMPORTANT.md injection tail",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; bimodal across 249 invocations: p50 43ms vs p95 1551ms vs max 2068ms (paths seen: PreToolUse, SessionStart, (unattributed)) — a cheap path and an expensive path with no cache between them",
+      "p50ms": null,
+      "p95ms": 1551,
+      "totalMs": 115792,
+      "invocations": 249,
+      "rootCause": "the renderer re-reads PROJECT_IMPORTANT.md and re-runs its full work (strict UTF-8 decode, control-character validation, sensitive-value scan, envelope render) on every invocation; there is no mtime/content-keyed cache, so repeat runs pay the whole cost instead of replaying the previous result",
+      "fixTask": "TASK-235",
+      "status": "confirmed"
+    },
+    {
+      "id": "record-execution-multiplicity",
+      "area": "record-execution.sh — PostToolUse receipt writer registered in 3 matcher groups",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; measured fire count per tool call = 12 (distribution {\"1\":4341,\"3\":1,\"4\":1,\"12\":1} over 4344 calls). The \"once per matching matcher group\" repeat hypothesis is REFUTED: settings.json names the hook in 3 groups, but those matchers are mutually exclusive per tool, so it runs exactly once per call. Its real cost is breadth — highest-invocation hook measured (n=4360, p50 39ms, total 186663ms).",
+      "p50ms": 39,
+      "p95ms": 64,
+      "totalMs": 186663,
+      "invocations": 4360,
+      "rootCause": "not a duplication bug: the matchers (Read|Grep|Glob, Edit|Write, Bash) are mutually exclusive, so at most one group fires per tool. The cost is that one full 3-spawn hook row is paid on every tool call of every kind — breadth, not repeats.",
+      "fixTask": null,
+      "status": "not-a-bug"
+    },
+    {
+      "id": "index-router-cold-start",
+      "area": "index/router helpers invoked per prompt (skill-router → resolve-context / route-task / query-index)",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; UserPromptSubmit groups: n=193, p50 319ms, p95 585ms, 4 hook rows/group. Standalone cold starts: skill-router.sh 89.6ms, query-index.mjs 24.1ms, route-task.mjs not benchmarked, resolve-context.mjs not benchmarked",
+      "p50ms": 319,
+      "p95ms": 585,
+      "totalMs": 65511,
+      "invocations": 193,
+      "rootCause": "every router helper pays a fresh node cold start (no warm process is reused), and skill-router.sh additionally imports the index core in-process per invocation; the cost is process boot plus module load, not the routing decision itself",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "c29-c32-hot-path-additions",
+      "area": "recent hot-path additions — 20 files across 16 commits",
+      "evidence": "git-log",
+      "evidenceDetail": "git log --since=14 days ago --diff-filter=A -- templates/.claude/hooks templates/.claude/ukit/runtime templates/.claude/ukit/index: templates/.claude/hooks/session-episode.sh, templates/.claude/hooks/project-important.sh, templates/.claude/hooks/project-important.sh, templates/.claude/ukit/runtime/project-important.mjs, templates/.claude/ukit/runtime/sensitive-value-scanner.mjs, templates/.claude/ukit/runtime/hook-chain-budget.mjs, templates/.claude/ukit/runtime/stop-coordinator.mjs, templates/.claude/ukit/runtime/hook-payload-store.mjs, templates/.claude/ukit/runtime/hook-telemetry.mjs, templates/.claude/ukit/runtime/hook-telemetry.sh, templates/.claude/ukit/runtime/context-capacity.mjs, templates/.claude/ukit/runtime/async-lock.mjs, +8 more",
+      "p50ms": null,
+      "p95ms": null,
+      "totalMs": null,
+      "invocations": 20,
+      "rootCause": "each added hook file becomes another settings.json row on the event it joins, i.e. another 3-spawn hook row per call (SPEC §1.3 consolidation is what removes them); each added runtime module is an extra cold-start import paid per invocation. The regression is additive spawn count, not one slow file.",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "existing-chain-runner-baseline",
+      "area": "hook-chain-runner.mjs — already-consolidated omp path (TASK-234 reuse target)",
+      "evidence": "telemetry",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; a whole chain of 1 scripts runs in ONE node process at p50 106ms / p95 409ms (n=1753), while a Claude settings group of comparable size pays 3 spawns per script. Nested per-script rows: record-execution.sh p50 71ms, sensitive-data-guard.sh p50 60ms, handoff-model-guard.sh p50 41ms, context-hardcap-gate.sh p50 72ms, auto-allow-bash.sh p50 72ms, block-dangerous.sh p50 82ms",
+      "p50ms": 106,
+      "p95ms": 409,
+      "totalMs": 283464,
+      "invocations": 1753,
+      "rootCause": "no defect — this is the measured proof that one-process-per-group execution is achievable with per-script telemetry intact, so TASK-234 should re-point settings.json at this runner rather than build a new one",
+      "fixTask": "TASK-234",
+      "status": "confirmed"
+    },
+    {
+      "id": "process-floor",
+      "area": "process spawn floor (bash + node + advisory telemetry child)",
+      "evidence": "micro-benchmark",
+      "evidenceDetail": "live spawnSync over 10 iterations: bash -c true p50 3.76ms, node -e 0 p50 16.94ms, node hook-telemetry.mjs --finish p50 19.53ms",
+      "p50ms": 40.235959,
+      "p95ms": 45.548333,
+      "totalMs": null,
+      "invocations": 0,
+      "rootCause": "each settings.json hook row is a bash wrapper that boots node, then arms a second node child for advisory telemetry; this floor is paid per row, so it multiplies by the row count of the event",
+      "fixTask": "TASK-234",
+      "status": "not-a-bug"
+    },
+    {
+      "id": "hook-reinject-context-sh-unmeasured",
+      "area": "single hook — reinject-context.sh (registered in settings.json, no telemetry rows)",
+      "evidence": "unavailable",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; no row exists for this script, so no per-invocation number is reported. It still costs 3 spawns on its event and is therefore already counted inside that chain's total.",
+      "p50ms": null,
+      "p95ms": null,
+      "totalMs": null,
+      "invocations": 0,
+      "rootCause": "no measurement exists for reinject-context.sh on this tree — its cost stays unset (evidenced by absence) rather than estimated; the chain it belongs to carries the spawn floor",
+      "fixTask": null,
+      "status": "deferred"
+    },
+    {
+      "id": "hook-auto-prune-bash-sh-unmeasured",
+      "area": "single hook — auto-prune-bash.sh (registered in settings.json, no telemetry rows)",
+      "evidence": "unavailable",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; no row exists for this script, so no per-invocation number is reported. It still costs 3 spawns on its event and is therefore already counted inside that chain's total.",
+      "p50ms": null,
+      "p95ms": null,
+      "totalMs": null,
+      "invocations": 0,
+      "rootCause": "no measurement exists for auto-prune-bash.sh on this tree — its cost stays unset (evidenced by absence) rather than estimated; the chain it belongs to carries the spawn floor",
+      "fixTask": null,
+      "status": "deferred"
+    },
+    {
+      "id": "hook-reset-compact-pressure-sh-unmeasured",
+      "area": "single hook — reset-compact-pressure.sh (registered in settings.json, no telemetry rows)",
+      "evidence": "unavailable",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; no row exists for this script, so no per-invocation number is reported. It still costs 3 spawns on its event and is therefore already counted inside that chain's total.",
+      "p50ms": null,
+      "p95ms": null,
+      "totalMs": null,
+      "invocations": 0,
+      "rootCause": "no measurement exists for reset-compact-pressure.sh on this tree — its cost stays unset (evidenced by absence) rather than estimated; the chain it belongs to carries the spawn floor",
+      "fixTask": null,
+      "status": "deferred"
+    },
+    {
+      "id": "hook-session-episode-sh-unmeasured",
+      "area": "single hook — session-episode.sh (registered in settings.json, no telemetry rows)",
+      "evidence": "unavailable",
+      "evidenceDetail": ".ukit/storage/cache/hook-latency — 34629 rows across 55 session files, span 2026-09-17T10:50:18.670Z → 2026-09-20T07:42:45.514Z; no row exists for this script, so no per-invocation number is reported. It still costs 3 spawns on its event and is therefore already counted inside that chain's total.",
+      "p50ms": null,
+      "p95ms": null,
+      "totalMs": null,
+      "invocations": 0,
+      "rootCause": "no measurement exists for session-episode.sh on this tree — its cost stays unset (evidenced by absence) rather than estimated; the chain it belongs to carries the spawn floor",
+      "fixTask": null,
+      "status": "deferred"
+    }
+  ]
+}

package/src/cli/adapters.js

@@ -19,16 +19,6 @@ export const OPTIONAL_ADAPTERS = [
       '.codex/agents',
     ],
   },
-  {
-    key: 'opencode',
-    label: 'OpenCode',
-    itemIds: [
-      'multi-opencode-config',
-    ],
-    managedPaths: [
-      'opencode.json',
-    ],
-  },
   {
     key: 'omp',
     label: 'Oh My Pi (omp)',
@@ -75,13 +65,11 @@ export const TOOL_ALIASES = new Map([
   ['claude-code', 'claude'],
   ['codex', 'codex'],
   ['openai-codex', 'codex'],
-  ['opencode', 'opencode'],
-  ['open-code', 'opencode'],
   ['omp', 'omp'],
   ['oh-my-pi', 'omp'],
 ]);
 
-export const TOOLS_FLAG_HELP = 'all|none|claude,codex,opencode,omp';
+export const TOOLS_FLAG_HELP = 'all|none|claude,codex,omp';
 
 export function parseToolsArg(argv = []) {
   let toolsArg = null;
@@ -137,6 +125,15 @@ export function resolveOptionalToolKeys(toolsArg) {
       );
     }
 
+    // legacy: 'opencode'/'open-code' were adapters until the OpenCode lane was removed.
+    // Warn + ignore so `--tools=opencode` still exits 0 instead of hard-failing.
+    if (normalizedToken === 'opencode' || normalizedToken === 'open-code') {
+      console.warn(
+        `[UKit] '${token}' was removed as a UKit adapter — ignoring. Allowed: ${TOOLS_FLAG_HELP}`,
+      );
+      continue;
+    }
+
     const normalizedTool = TOOL_ALIASES.get(normalizedToken) ?? null;
     if (!normalizedTool) {
       throw new Error(`Unknown tool '${token}'. Allowed: ${TOOLS_FLAG_HELP}`);

package/src/cli/commands/doctor.js

@@ -24,10 +24,12 @@ import {
 } from '../../core/projectImportant.js';
 import { runDocContractChecks } from '../../core/docContracts.js';
 import { inspectUnattendedMode } from '../../core/unattendedDoctor.js';
+import { inspectPermissions } from '../../core/permissionDoctor.js';
+import { findOpencodeArtifacts, opencodeSteerMessage } from '../../core/opencodeSteer.js';
 
 export const DOCTOR_HELP_FLAGS = new Set(['--help', '-h']);
-const KNOWN_FLAGS = new Set([...DOCTOR_HELP_FLAGS, '--skills', '--gateway', '--docs']);
-const SUPPORTED_FLAGS_LIST = '--help, -h, --skills, --gateway, --docs';
+const KNOWN_FLAGS = new Set([...DOCTOR_HELP_FLAGS, '--skills', '--gateway', '--docs', '--permissions', '--json']);
+const SUPPORTED_FLAGS_LIST = '--help, -h, --skills, --gateway, --docs, --permissions, --json';
 
 export function printDoctorHelp() {
   console.log('Usage: ukit doctor [options]');
@@ -39,6 +41,30 @@ export function printDoctorHelp() {
   console.log('  --skills        Also print a skill word-count/budget report');
   console.log('  --gateway       Live gateway probe (streaming + non-streaming); advisory only');
   console.log('  --docs          Run doc-contract checks (manifests/documentation.yaml projects only)');
+  console.log('  --permissions   Print the full per-host permission report table');
+  console.log('  --json          Emit the permission report as JSON ({permissions:{...}})');
+}
+
+function printPermissionSection(permissionReport, { verbose }) {
+  console.log('[UKit] Permission report:');
+  console.log(`[UKit]   declared: ${permissionReport.declared} → effective: ${permissionReport.effective} (allowUnsafeEval=${permissionReport.allowUnsafeEval})`);
+  for (const [host, info] of Object.entries(permissionReport.hosts)) {
+    if (!info.present) {
+      if (verbose) console.log(`[UKit]   ${host}: SKIP (no UKit-managed surface)`);
+      continue;
+    }
+    console.log(`[UKit]   ${host}: ${info.enforcement} | prompts=${info.promptCount} | deny=${info.denyCount} (missing ${info.denyMissing.length}) | eval=${info.evalSurface ?? info.evalPosture ?? 'n/a'}`);
+  }
+  for (const check of permissionReport.checks) {
+    if (check.applicable === false) {
+      if (verbose) console.log(`[UKit]   -  ${check.label} (skip: ${check.detail ?? 'not applicable'})`);
+      continue;
+    }
+    const sev = check.severity === 'warning' ? ' (warning)' : check.severity === 'info' ? ' (info)' : '';
+    console.log(`[UKit]   ${check.passed ? '✓' : '✗'}  ${check.label}${sev}`);
+    if (check.detail) console.log(`[UKit]      detail: ${check.detail}`);
+    if (!check.passed && check.remedy) console.log(`[UKit]      remedy: ${check.remedy}`);
+  }
 }
 
 export async function runDoctor({ packageRoot, projectRoot, argv = [], homeDir = os.homedir(), ompPath = 'omp' }) {
@@ -52,6 +78,15 @@ export async function runDoctor({ packageRoot, projectRoot, argv = [], homeDir =
     return;
   }
 
+  // TASK-006 / FR-006 — `--json` is a machine surface: emit ONLY the permission
+  // report object, exit 1 when any FAIL class fires.
+  if (argv.includes('--json')) {
+    const permissionReport = await inspectPermissions({ projectRoot, ompPath });
+    console.log(JSON.stringify({ permissions: permissionReport }, null, 2));
+    if (permissionReport.failures.length > 0) process.exitCode = 1;
+    return;
+  }
+
   const pathConfig = buildPathConfig({ packageRoot, projectRoot });
   const runtimePaths = buildRuntimePaths(projectRoot);
   const manifest = await loadManifest(pathConfig.manifestPath);
@@ -203,13 +238,16 @@ export async function runDoctor({ packageRoot, projectRoot, argv = [], homeDir =
       remedy: INSTALL_REPAIR,
     });
   }
-  if (piWiring.opencode.installed) {
+  // SPEC FR-007 — a leftover opencode.json is a WARN row: advisory only, never a
+  // FAIL, never auto-fixed. Removal is the user's manual action.
+  const opencodeArtifacts = findOpencodeArtifacts(projectRoot);
+  if (opencodeArtifacts.length > 0) {
     projectChecks.push({
-      label: 'OpenCode fallback present (AGENTS.md owner-instructions section)',
-      passed: piWiring.opencode.wired,
-      failed: !piWiring.opencode.wired,
-      remediationClass: 'install-repairable',
-      remedy: INSTALL_REPAIR,
+      label: opencodeSteerMessage(opencodeArtifacts),
+      passed: true,
+      severity: 'warning',
+      remediationClass: 'advisory',
+      remedy: 'Remove it manually: rm opencode.json',
     });
   }
 
@@ -252,7 +290,11 @@ export async function runDoctor({ packageRoot, projectRoot, argv = [], homeDir =
   console.log('[UKit] Project rules checks:');
   for (const check of projectChecks) {
     if (check.applicable === false) continue;
-    console.log(`[UKit]   ${ok(check.passed)}  ${check.label}`);
+    const severity = check.severity === 'warning' ? ' (warning)' : '';
+    console.log(`[UKit]   ${ok(check.passed)}  ${check.label}${severity}`);
+    if (check.remedy && check.severity === 'warning') {
+      console.log(`[UKit]      remedy: ${check.remedy}`);
+    }
   }
 
   // TASK-011 / SPEC §8 (FR-007): unattended-mode checks — same remediation-class
@@ -269,6 +311,11 @@ export async function runDoctor({ packageRoot, projectRoot, argv = [], homeDir =
     }
   }
 
+  // TASK-006 / FR-006 — unified permission report. Compact section by default,
+  // full per-host table under --permissions. FAIL classes block the exit code.
+  const permissionReport = await inspectPermissions({ projectRoot, ompPath });
+  printPermissionSection(permissionReport, { verbose: argv.includes('--permissions') });
+
   if (runtimeConfigInspection.errors.length > 0) {
     console.log(`[UKit] Runtime config issues: ${runtimeConfigInspection.errors.join(' | ')}`);
   }
@@ -420,7 +467,10 @@ export async function runDoctor({ packageRoot, projectRoot, argv = [], homeDir =
   const failedUnattendedChecks = unattended.checks.filter(
     (check) => check.applicable !== false && !check.passed,
   );
-  const allFailedChecks = [...failedProjectChecks, ...failedUnattendedChecks];
+  const failedPermissionChecks = permissionReport.checks.filter(
+    (check) => check.applicable !== false && !check.passed && check.severity !== 'info' && check.severity !== 'warning',
+  );
+  const allFailedChecks = [...failedProjectChecks, ...failedUnattendedChecks, ...failedPermissionChecks];
   const blockingFailures = allFailedChecks.filter(
     (check) => check.remediationClass === 'install-repairable' || check.remediationClass === 'owner-action',
   );

package/src/cli/index.js

@@ -117,8 +117,8 @@ export async function runCli({ argv, packageRoot, projectRoot, packageVersion })
     console.log('');
     console.log('Options (install / diff):');
     console.log('  --tools=<list>     Comma-separated adapters to enable (default: all)');
-    console.log('                     Values: claude, codex, opencode, omp');
-    console.log('                     Example: ukit install --tools=claude,codex,opencode,omp');
+    console.log('                     Values: claude, codex, omp');
+    console.log('                     Example: ukit install --tools=claude,codex,omp');
     console.log('                     Use --tools=none to disable all optional adapters');
     console.log('  --with-codegraph   Enable CodeGraph MCP integration in CLAUDE.md');
     console.log('                     Adds routing rules that prefer codegraph_* tools for');