harmoniqs · aarontrowbridge · Sep 13, 2026 · Sep 13, 2026 · Sep 13, 2026 · Sep 13, 2026
diff --git a/AGENTS.md b/AGENTS.md
@@ -27,7 +27,7 @@ Interview flows are data (`packages/extension/scores/`). Schemas: `packages/sche
 ## Setup (in order; each step has a check)
 
 ```bash
-git clone git@github.com:harmoniqs/amicissimode.git && cd amicode
+git clone git@github.com:harmoniqs/amicode.git && cd amicode
 corepack enable && pnpm install               # check: exits 0, lockfile untouched
 pnpm -r build                                 # check: packages/extension/dist/extension.js exists
 pnpm --filter amicode run fetch:opencode   # check: vendor/opencode/<platform>/opencode exists

diff --git a/packages/extension/agents/develop.md b/packages/extension/agents/develop.md
@@ -90,14 +90,19 @@ record; this prose is the binding, never a second spec.
   (when human-in-the-loop) is by a reviewer who is never the implementer.
   **No per-slice PRs in orchestrated mode.** After each slice's branch merges
   into the integration branch, **remove its worktree immediately** via
-  `amicode_workspace` — do not accumulate worktrees across slices.
+  `amicode_workspace` — do not accumulate worktrees across slices. The gate
+  suite includes the skills-integrity lint when present on the branch, and
+  merged work that changed a package API surface is grep-checked against
+  skill references — hits become ledger findings.
 
 **Dispatch discipline** — the implementer (the `implementer` subagent card) is
 your only writer role: fresh context per slice, one issue per cast, worktree-
 bound, no PR and no merge in orchestrated mode. You gate and merge; the
 implementer returns the branch and a structured result (issue, status,
 branch, commit_shas, ac_results, notes). A slice that hits its step limit
-returns `EXHAUSTED:` — an open loop for you, not a failure to hide.
+returns `EXHAUSTED:` — an open loop for you, not a failure to hide. The
+loop-boundary ledger update includes the campaign's skill delta — findings
+filed, skills touched, proposals pending.
 
 **Hard rules** — issue + PR for all package work (the development gate); never
 merge non-green work; never push directly to protected branches; promotion

diff --git a/packages/extension/agents/research.md b/packages/extension/agents/research.md
@@ -75,7 +75,11 @@ run the spec review (`--allow-unreviewed` forbidden for launch-shaped work)
 experiment, assigned env per the checkout registry) → run the gates YOURSELF
 via bash (verdicts derived from commands, never self-reported; no LLM judges
 a fidelity claim) → cast the **analyzer** (read-only, raw artifacts only) →
-commit the ledger update → repeat.
+commit the ledger update → repeat. The analyzer cast carries the
+skill-friction check — instructions-vs-reality mismatches become findings,
+repeated un-skilled idioms become skill proposals — and the record boundary
+carries the same skill-delta row (findings filed, skills touched, proposals
+pending).
 
 **The boundary** — read-only ops and bounded side-effect-free one-liners are
 probes; the moment work writes a durable file, launches a solve, or runs a

diff --git a/packages/extension/modes/develop/card.md b/packages/extension/modes/develop/card.md
@@ -90,14 +90,19 @@ record; this prose is the binding, never a second spec.
   (when human-in-the-loop) is by a reviewer who is never the implementer.
   **No per-slice PRs in orchestrated mode.** After each slice's branch merges
   into the integration branch, **remove its worktree immediately** via
-  `amicode_workspace` — do not accumulate worktrees across slices.
+  `amicode_workspace` — do not accumulate worktrees across slices. The gate
+  suite includes the skills-integrity lint when present on the branch, and
+  merged work that changed a package API surface is grep-checked against
+  skill references — hits become ledger findings.
 
 **Dispatch discipline** — the implementer (the `implementer` subagent card) is
 your only writer role: fresh context per slice, one issue per cast, worktree-
 bound, no PR and no merge in orchestrated mode. You gate and merge; the
 implementer returns the branch and a structured result (issue, status,
 branch, commit_shas, ac_results, notes). A slice that hits its step limit
-returns `EXHAUSTED:` — an open loop for you, not a failure to hide.
+returns `EXHAUSTED:` — an open loop for you, not a failure to hide. The
+loop-boundary ledger update includes the campaign's skill delta — findings
+filed, skills touched, proposals pending.
 
 **Hard rules** — issue + PR for all package work (the development gate); never
 merge non-green work; never push directly to protected branches; promotion

diff --git a/packages/extension/modes/research/card.md b/packages/extension/modes/research/card.md
@@ -75,7 +75,11 @@ run the spec review (`--allow-unreviewed` forbidden for launch-shaped work)
 experiment, assigned env per the checkout registry) → run the gates YOURSELF
 via bash (verdicts derived from commands, never self-reported; no LLM judges
 a fidelity claim) → cast the **analyzer** (read-only, raw artifacts only) →
-commit the ledger update → repeat.
+commit the ledger update → repeat. The analyzer cast carries the
+skill-friction check — instructions-vs-reality mismatches become findings,
+repeated un-skilled idioms become skill proposals — and the record boundary
+carries the same skill-delta row (findings filed, skills touched, proposals
+pending).
 
 **The boundary** — read-only ops and bounded side-effect-free one-liners are
 probes; the moment work writes a durable file, launches a solve, or runs a

diff --git a/packages/extension/skills/research/SKILL.md b/packages/extension/skills/research/SKILL.md
@@ -104,7 +104,7 @@ running `amico env promote` on any staged proposals in `ledger/environment-propo
 
 ## The campaign ledger (create at kickoff, before any work)
 
-Path: `<project>/ledger/campaigns/campaign-<YYYYMMDD>-<slug>.md`. Nine sections, in order:
+Path: `<project>/ledger/campaigns/campaign-<YYYYMMDD>-<slug>.md` — this project-local ledger serves Research-Project-scoped work and reconciles to the **session ledger** in the personal vault (`sessions/session-<YYYYMMDD>-<slug>.md`, per `director-core`'s discovery rule) at each loop boundary. Nine sections, in order:
 
 1. Objective & standing directives
 2. Hypothesis ledger (H# → verdict → evidence → wiki-links)

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/confidence-wrong-type.json b/packages/extension/test/fixtures/vault-cards/invalid/confidence-wrong-type.json
@@ -2,6 +2,7 @@
   "confidence": 5,
   "date": "2026-08-22",
   "evidence": [],
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/experiment-fidelity-negative.json b/packages/extension/test/fixtures/vault-cards/invalid/experiment-fidelity-negative.json
@@ -5,8 +5,8 @@
   "gate": "X",
   "platform": "transmon",
   "session_id": "s",
-  "status": "solved",
+  "status": "completed",
   "tags": [],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/invalid/experiment-fidelity-out-of-range.json b/packages/extension/test/fixtures/vault-cards/invalid/experiment-fidelity-out-of-range.json
@@ -5,8 +5,8 @@
   "gate": "X",
   "platform": "transmon",
   "session_id": "s",
-  "status": "solved",
+  "status": "completed",
   "tags": [],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/invalid/experiment-missing-fidelity.json b/packages/extension/test/fixtures/vault-cards/invalid/experiment-missing-fidelity.json
@@ -4,8 +4,8 @@
   "gate": "X",
   "platform": "transmon",
   "session_id": "s",
-  "status": "solved",
+  "status": "completed",
   "tags": [],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/invalid/experiment-negative-duration.json b/packages/extension/test/fixtures/vault-cards/invalid/experiment-negative-duration.json
@@ -5,8 +5,8 @@
   "gate": "X",
   "platform": "transmon",
   "session_id": "s",
-  "status": "solved",
+  "status": "completed",
   "tags": [],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/invalid/malformed-review-by.json b/packages/extension/test/fixtures/vault-cards/invalid/malformed-review-by.json
@@ -3,6 +3,7 @@
   "date": "2026-08-22",
   "evidence": [],
   "review_by": "last tuesday",
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/provenance-not-array.json b/packages/extension/test/fixtures/vault-cards/invalid/provenance-not-array.json
@@ -3,6 +3,7 @@
   "date": "2026-08-22",
   "evidence": [],
   "provenance": "not-an-array",
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/review-by-calendar-invalid.json b/packages/extension/test/fixtures/vault-cards/invalid/review-by-calendar-invalid.json
@@ -3,6 +3,7 @@
   "date": "2026-08-22",
   "evidence": [],
   "review_by": "2026-02-30",
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/review-by-wrong-type.json b/packages/extension/test/fixtures/vault-cards/invalid/review-by-wrong-type.json
@@ -3,6 +3,7 @@
   "date": "2026-08-22",
   "evidence": [],
   "review_by": 2026,
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/sentinel-conflict.json b/packages/extension/test/fixtures/vault-cards/invalid/sentinel-conflict.json
@@ -7,6 +7,7 @@
   ],
   "provenance_unrecoverable": true,
   "reviewed_after": "journal/pass-1.md",
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/sentinel-empty-reviewed-after.json b/packages/extension/test/fixtures/vault-cards/invalid/sentinel-empty-reviewed-after.json
@@ -5,6 +5,7 @@
   "provenance": [],
   "provenance_unrecoverable": true,
   "reviewed_after": "",
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/sentinel-no-review-pointer.json b/packages/extension/test/fixtures/vault-cards/invalid/sentinel-no-review-pointer.json
@@ -3,6 +3,7 @@
   "date": "2026-08-22",
   "evidence": [],
   "provenance_unrecoverable": true,
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/invalid/unknown-confidence.json b/packages/extension/test/fixtures/vault-cards/invalid/unknown-confidence.json
@@ -2,6 +2,7 @@
   "confidence": "certain",
   "date": "2026-08-22",
   "evidence": [],
+  "session_id": null,
   "source": "s",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/valid-md/insight-cat-state.md b/packages/extension/test/fixtures/vault-cards/valid-md/insight-cat-state.md
@@ -1,6 +1,7 @@
 ---
 type: insight
 date: 2026-08-22
+session_id: 20260913-cat-state
 source: session-2026-08-21
 evidence:
   - experiments/e-cat.md

diff --git a/packages/extension/test/fixtures/vault-cards/valid/device-mini.json b/packages/extension/test/fixtures/vault-cards/valid/device-mini.json
@@ -1,4 +1,6 @@
 {
+  "device_class": "classical",
+  "location": "desk",
   "name": "Mac mini",
   "platforms": [
     "server"

diff --git a/packages/extension/test/fixtures/vault-cards/valid/experiment-cavity-failed.json b/packages/extension/test/fixtures/vault-cards/valid/experiment-cavity-failed.json
@@ -9,6 +9,6 @@
   "tags": [
     "fock"
   ],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/valid/experiment-fluxonium-ext.json b/packages/extension/test/fixtures/vault-cards/valid/experiment-fluxonium-ext.json
@@ -9,11 +9,11 @@
   ],
   "review_by": "2026-11-16",
   "session_id": "s4",
-  "status": "solved",
+  "status": "completed",
   "subject": "fluxonium-x",
   "tags": [
     "fluxonium"
   ],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/valid/experiment-rydberg-cz.json b/packages/extension/test/fixtures/vault-cards/valid/experiment-rydberg-cz.json
@@ -5,10 +5,10 @@
   "gate": "CZ",
   "platform": "rydberg",
   "session_id": "s2",
-  "status": "solved",
+  "status": "completed",
   "tags": [
     "cz"
   ],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/valid/experiment-transmon-x.json b/packages/extension/test/fixtures/vault-cards/valid/experiment-transmon-x.json
@@ -5,10 +5,10 @@
   "gate": "X",
   "platform": "transmon",
   "session_id": "s1",
-  "status": "solved",
+  "status": "completed",
   "tags": [
     "x-gate"
   ],
-  "task_type": "experiment-sim",
+  "task_type": "experiment",
   "type": "experiment"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/valid/hypothesis-falsified.json b/packages/extension/test/fixtures/vault-cards/valid/hypothesis-falsified.json
@@ -3,8 +3,9 @@
   "evidence": [
     "experiments/e6.md"
   ],
+  "session_id": null,
   "source": "session",
-  "status": "falsified",
+  "status": "refuted",
   "tags": [
     "magnus"
   ],

diff --git a/packages/extension/test/fixtures/vault-cards/valid/hypothesis-open.json b/packages/extension/test/fixtures/vault-cards/valid/hypothesis-open.json
@@ -1,6 +1,7 @@
 {
   "date": "2026-08-18",
   "evidence": [],
+  "session_id": null,
   "source": "session",
   "status": "open",
   "tags": [

diff --git a/packages/extension/test/fixtures/vault-cards/valid/hypothesis-tested.json b/packages/extension/test/fixtures/vault-cards/valid/hypothesis-tested.json
@@ -3,8 +3,9 @@
   "evidence": [
     "experiments/e5.md"
   ],
+  "session_id": null,
   "source": "session",
-  "status": "tested",
+  "status": "confirmed",
   "tags": [
     "bilinear"
   ],

diff --git a/packages/extension/test/fixtures/vault-cards/valid/insight-full-extensions.json b/packages/extension/test/fixtures/vault-cards/valid/insight-full-extensions.json
@@ -8,6 +8,7 @@
     "experiments/e2.md"
   ],
   "review_by": "2026-11-18",
+  "session_id": null,
   "source": "session",
   "subject": "warm-starts",
   "tags": [

diff --git a/packages/extension/test/fixtures/vault-cards/valid/insight-minimal.json b/packages/extension/test/fixtures/vault-cards/valid/insight-minimal.json
@@ -2,6 +2,7 @@
   "confidence": "medium",
   "date": "2026-08-19",
   "evidence": [],
+  "session_id": null,
   "source": "retro",
   "tags": [],
   "type": "insight"

diff --git a/packages/extension/test/fixtures/vault-cards/valid/insight-plain.json b/packages/extension/test/fixtures/vault-cards/valid/insight-plain.json
@@ -4,6 +4,7 @@
   "evidence": [
     "experiments/e1.md"
   ],
+  "session_id": null,
   "source": "session",
   "tags": [
     "warm-starts"

diff --git a/packages/extension/test/fixtures/vault-cards/valid/insight-routing.json b/packages/extension/test/fixtures/vault-cards/valid/insight-routing.json
@@ -4,6 +4,7 @@
   "evidence": [
     "experiments/e4.md"
   ],
+  "session_id": null,
   "source": "session",
   "subject": "model-routing",
   "tags": [

diff --git a/packages/extension/test/fixtures/vault-cards/valid/insight-sentinel-reviewed.json b/packages/extension/test/fixtures/vault-cards/valid/insight-sentinel-reviewed.json
@@ -7,6 +7,7 @@
   "provenance": [],
   "provenance_unrecoverable": true,
   "reviewed_after": "journal/pass-1.md",
+  "session_id": null,
   "source": "session",
   "tags": [
     "legacy"

diff --git a/packages/extension/test/fixtures/vault-cards/valid/meeting-design.json b/packages/extension/test/fixtures/vault-cards/valid/meeting-design.json
@@ -3,8 +3,10 @@
     "aaron"
   ],
   "date": "2026-08-18",
+  "org": "[[org-harmoniqs]]",
   "tags": [
     "design"
   ],
+  "topic": "weekly sync",
   "type": "meeting"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/valid/meeting-weekly.json b/packages/extension/test/fixtures/vault-cards/valid/meeting-weekly.json
@@ -4,8 +4,10 @@
     "kate"
   ],
   "date": "2026-08-15",
+  "org": "[[org-harmoniqs]]",
   "tags": [
     "weekly"
   ],
+  "topic": "weekly sync",
   "type": "meeting"
 }
diff --git a/packages/extension/test/fixtures/vault-cards/valid/method-bangbang.json b/packages/extension/test/fixtures/vault-cards/valid/method-bangbang.json
@@ -1,7 +1,10 @@
 {
-  "applicability": "short-time gates, drive-limited regimes",
+  "applicability": [
+    "short-time gates, drive-limited regimes"
+  ],
   "date": "2026-08-01",
   "name": "bang-bang parameterization",
+  "session_id": null,
   "source": "literature",
   "tags": [
     "parameterization"

diff --git a/packages/extension/test/fixtures/vault-cards/valid/method-smooth.json b/packages/extension/test/fixtures/vault-cards/valid/method-smooth.json
@@ -1,10 +1,13 @@
 {
-  "applicability": "transmon single-qubit gates",
+  "applicability": [
+    "transmon single-qubit gates"
+  ],
   "date": "2026-08-02",
   "name": "smooth parameterization",
   "provenance": [
     "experiments/e7.md"
   ],
+  "session_id": null,
   "source": "demo",
   "tags": [
     "parameterization"

diff --git a/packages/extension/test/fixtures/vault-cards/valid/org-harmoniqs.json b/packages/extension/test/fixtures/vault-cards/valid/org-harmoniqs.json
@@ -1,5 +1,7 @@
 {
+  "domain": "quantum control",
   "name": "Harmoniqs",
+  "relationship": "partner",
   "tags": [
     "team"
   ],