@dzhechkov/skills-feature-adr 1.3.65 → 1.3.67

package/.dz-manifest.json

@@ -13,7 +13,7 @@
       },
       {
         "path": "README.md",
-        "sha256": "4a887b8fe2a52c75be7189ddeef4393dc2a8a872d747bdbde3caba94488960dd"
+        "sha256": "1d123dcb15ac44b9adfe8266cb72d2dc69d5086f54003540795296c36e06e1b2"
       },
       {
         "path": "bin/cli.js",
@@ -25,7 +25,7 @@
       },
       {
         "path": "package.json",
-        "sha256": "1f95669ddfb76990bfeb0e1be603c27651f47a9233020ed10c37e8705543b688"
+        "sha256": "31de034ec7534b585ee99a2dac0105bc4e43dca0c5b419076e66c43e797509d3"
       },
       {
         "path": "scripts/sync-templates.js",
@@ -121,7 +121,7 @@
       },
       {
         "path": "templates/.claude/skills/feature-adr/SKILL.md",
-        "sha256": "c422d1e2b50cf7b8bd56cd5d886d96af30d8bf33a468b925396f395ebf32de53"
+        "sha256": "bdedfed69d985eb6969f158a72d92f63c6b1b993008c3661f7ffc2c6a7ccfbae"
       },
       {
         "path": "templates/.claude/skills/feature-adr/examples/sample-feature-output.md",
@@ -133,31 +133,31 @@
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/01-requirements.md",
-        "sha256": "7b599efd3122be7ce2247bba1c9204e1b3c09322ad56ed4cedcbd4dd471ed41e"
+        "sha256": "e6ea3a224b555d5b7410cf2ddf02a3fb8d009fa01109463353e2b993c318270c"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/02-research.md",
-        "sha256": "55099189082d872d08a9d40b11782d8d08256955cbea617aa457756eb9fa1a07"
+        "sha256": "e2506e42251c74485188473acc4586f841afbec3ae7c5eaa8a3d8f73f6e14cc5"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/03-adr.md",
-        "sha256": "742f1d362fd4d1b1b731fe81b7aa794886aa58126005ad67ef0a27775c80f082"
+        "sha256": "ea03eee136c89267888b8034baa1a1cb0057314a2e745df71654a9ebee5d46b6"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/03.5-ideation-swarm.md",
-        "sha256": "220e06d0d67643f09df868df32d1ec3acdc205653a5ef977d171791d9ce77b74"
+        "sha256": "d48a8d6eebf9760cc8f1a4f19df0e6afc31fcc1284b5eacfb604ad8ced6a0b53"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/04-ddd.md",
-        "sha256": "c6cf98ab0dd80183ad507aca12edf164ee79a5fb60eee84d674cf73fe3a68646"
+        "sha256": "e9504cf1183bc0105167dc06108860882941fcf1b5be79d1978d8212c410ffbb"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/05-architecture.md",
-        "sha256": "227ec469154348afc414ef4751cc34ea5bb87ecb063cf196eff254615a6b5e00"
+        "sha256": "2ef7cf316903f0cbb1fabe1f0e1770c22b2cb3a11a4118cac0529c25074b5e8a"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/06-implementation-plan.md",
-        "sha256": "7a6c830d63f73953a96f1cb125ab88b7d5c24b2d4b61e997d6d3dfea8262d495"
+        "sha256": "c8fdd1877d8d9515ab29f52c3db49ddd9612b5797bc071e6208def860c02e6f2"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/07-code.md",
@@ -165,7 +165,7 @@
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/08-qe.md",
-        "sha256": "0d314eaeb69a761f68c4beb1ab7df0a2da1964538a8eb33c99cb9b819598d255"
+        "sha256": "97e9a7bc54f196b381c4ef71258cddcb2fb48f216df1ba015927c176bad57084"
       },
       {
         "path": "templates/.claude/skills/feature-adr/modules/09-fleet-qe.md",
@@ -255,6 +255,10 @@
         "path": "templates/.claude/skills/feature-adr/references/qe-checklist.md",
         "sha256": "238d8896996dc53559f58b24aa7eca966cc8845c3b9dda6d982c717c657b91e4"
       },
+      {
+        "path": "templates/.claude/skills/feature-adr/scripts/check-plan-completeness.mjs",
+        "sha256": "41d85b012a9d6cc172eaa5d7777294f2c01fb54b752ed61c7cbe2d2cd7b79b67"
+      },
       {
         "path": "templates/.claude/skills/frontend-design/LICENSE.txt",
         "sha256": "0d542e0c8804e39aa7f37eb00da5a762149dc682d7829451287e11b938e94594"
@@ -317,7 +321,7 @@
       },
       {
         "path": "templates/.claude/workflows/feature-adr.js",
-        "sha256": "8afa77bb421ea4052eec0feab5f49edc1f2b1cea69238971f8ef1b3c8328a483"
+        "sha256": "d9cda8df501188c894d083c5c44c240c9ebe9ba7e2d418302c448d73b089e051"
       },
       {
         "path": "templates/lib/memory-protocol.md",
@@ -345,5 +349,5 @@
       }
     ]
   },
-  "signature": "3OT+p3Tm5tBnn9T2RKEPlCnhCiaemKAOraeoPA4zYR28p1MpACnkeEmVyXPpwnqhG/a6uD/AjC/DrRuihMdHCw=="
+  "signature": "ZnkOOhYGfaIEp1W5L/Gphs0uniT0FieL/yn5KPl6cZ3zQeXkZcqDjfOta9b6zmPEgztZAclspZ0yf82VnjeHBQ=="
 }

package/README.md

@@ -531,6 +531,15 @@ gate > reviewer judgment > memory). A miss usually means a check lived one layer
   control that proved the path was I/O-free. The bundled role-default skills gained the matching entries
   (`code-impl` P25, `code-critic` AP14).
 
+- **No-stubs gate** (backlog 0b403a0106103901, Karpathy-Michaels rule XI). Step 8 greps the files THE RUN
+  touched for unfinished-stub markers — `TODO` / `FIXME` / `HACK` / `XXX` / `PLACEHOLDER` (case-sensitive,
+  word-bounded: `hackathon`/`todos` never fire) plus the `implement later` phrase — **any unwaived match =
+  the task shipped incomplete** (HIGH gap naming file:line). A line may carry an inline
+  `no-stubs: <reason>` waiver; a waiver **without** a reason is itself a HIGH gap, never an exemption. The
+  same scan runs mechanically at publish time as the SOFT `no-stubs` rule in `dz guard check --op publish`
+  (change-set scoped — a deliberate design: a tree-wide scan measured ~78% ancient-marker noise). Markers
+  quoted in the QE report are backticked so the report itself scans clean.
+
 Both forms carry all of it: the interactive skill (step modules + banner template) and the deterministic
 workflow (stage prompts + derived `gates` in its returns) — same shapes, same vocabulary.
 
@@ -869,3 +878,38 @@ them inside the package's own monorepo is almost never what you want anyway.
 - **npm:** [https://www.npmjs.com/package/@dzhechkov/skills-feature-adr](https://www.npmjs.com/package/@dzhechkov/skills-feature-adr)
 - **Keysarium:** [https://www.npmjs.com/package/@dzhechkov/keysarium](https://www.npmjs.com/package/@dzhechkov/keysarium)
 - **BTO:** [https://www.npmjs.com/package/@dzhechkov/skills-bto](https://www.npmjs.com/package/@dzhechkov/skills-bto)
+
+## Write discipline — why the writing steps now build a skeleton first
+
+MEASURED in the field on 2026-08-19 and 2026-08-20, two independent runs of the canonical pipeline
+against a ~130-file repository: the steps that must produce a document — Step 5 (Architecture) and
+Step 6 (Plan) — **never reached a write**. 18 attempts, zero file writes in every one. They died in
+the READING phase: a shell result returns, the agent thinks about its next move, stays silent past
+the runtime's 180-second inactivity watchdog, and is killed. One run cost ~4M tokens and 1h54m.
+
+The failure is **deterministic, not unlucky**: thinking time grows with accumulated history, so on a
+large enough repository unbounded exploration guarantees the kill. And the cause was in the
+instructions, not the runtime — every writing step said, in effect, *"read the code, write the
+document"*, with no reading budget and no order of operations. An agent obeying literally reads
+until it dies.
+
+The owner's control experiment is the whole evidence base, and it is n=1: same slice, same inputs,
+same model, ONE added paragraph about write discipline → a 10-section skeleton on disk 8 minutes in,
+first attempt, after six consecutive deaths.
+
+So every document-producing step now carries:
+
+1. the FIRST file write happens within the first ~12 tool calls — a skeleton of section headings
+   with one line of intent under each;
+2. then fill it **one section per edit**, no edit longer than ~120 lines;
+3. never go more than 2 minutes without a tool call;
+4. when unsure whether to read more or to write — **write**.
+
+**What this does not do**, stated plainly: it does not stop the deaths. The watchdog is unchanged.
+It changes what survives one — previously nothing, now a skeleton on disk that the next attempt can
+continue. The `~12` and `~120` are chosen parameters, not measured optima. Two possible regressions
+were named by an independent reviewer and are **not measured**: an early skeleton may anchor a
+structure chosen before understanding, and section-per-edit raises tool-call overhead.
+
+Steps whose deliverable is a returned verdict rather than a document (the complexity router) are
+deliberately excluded — "skeleton first" is nonsense there.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@dzhechkov/skills-feature-adr",
-  "version": "1.3.65",
+  "version": "1.3.67",
   "description": "Adaptive Feature Development skill pack for Claude Code — 11-step pipeline with Complexity Router (S/M/L/XL), ADR-driven architecture, 15 agentic-qe skills, multi-agent fleet QE. Supports --full-qe, --full-qe-extended, --with-learning, and --knowledge-extractor modes.",
   "bin": {
     "skills-feature-adr": "./bin/cli.js"

package/sbom.json

@@ -35,7 +35,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "4a887b8fe2a52c75be7189ddeef4393dc2a8a872d747bdbde3caba94488960dd"
+          "content": "1d123dcb15ac44b9adfe8266cb72d2dc69d5086f54003540795296c36e06e1b2"
         }
       ]
     },
@@ -65,7 +65,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "1f95669ddfb76990bfeb0e1be603c27651f47a9233020ed10c37e8705543b688"
+          "content": "31de034ec7534b585ee99a2dac0105bc4e43dca0c5b419076e66c43e797509d3"
         }
       ]
     },
@@ -305,7 +305,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "c422d1e2b50cf7b8bd56cd5d886d96af30d8bf33a468b925396f395ebf32de53"
+          "content": "bdedfed69d985eb6969f158a72d92f63c6b1b993008c3661f7ffc2c6a7ccfbae"
         }
       ]
     },
@@ -335,7 +335,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "7b599efd3122be7ce2247bba1c9204e1b3c09322ad56ed4cedcbd4dd471ed41e"
+          "content": "e6ea3a224b555d5b7410cf2ddf02a3fb8d009fa01109463353e2b993c318270c"
         }
       ]
     },
@@ -345,7 +345,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "55099189082d872d08a9d40b11782d8d08256955cbea617aa457756eb9fa1a07"
+          "content": "e2506e42251c74485188473acc4586f841afbec3ae7c5eaa8a3d8f73f6e14cc5"
         }
       ]
     },
@@ -355,7 +355,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "742f1d362fd4d1b1b731fe81b7aa794886aa58126005ad67ef0a27775c80f082"
+          "content": "ea03eee136c89267888b8034baa1a1cb0057314a2e745df71654a9ebee5d46b6"
         }
       ]
     },
@@ -365,7 +365,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "220e06d0d67643f09df868df32d1ec3acdc205653a5ef977d171791d9ce77b74"
+          "content": "d48a8d6eebf9760cc8f1a4f19df0e6afc31fcc1284b5eacfb604ad8ced6a0b53"
         }
       ]
     },
@@ -375,7 +375,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "c6cf98ab0dd80183ad507aca12edf164ee79a5fb60eee84d674cf73fe3a68646"
+          "content": "e9504cf1183bc0105167dc06108860882941fcf1b5be79d1978d8212c410ffbb"
         }
       ]
     },
@@ -385,7 +385,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "227ec469154348afc414ef4751cc34ea5bb87ecb063cf196eff254615a6b5e00"
+          "content": "2ef7cf316903f0cbb1fabe1f0e1770c22b2cb3a11a4118cac0529c25074b5e8a"
         }
       ]
     },
@@ -395,7 +395,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "7a6c830d63f73953a96f1cb125ab88b7d5c24b2d4b61e997d6d3dfea8262d495"
+          "content": "c8fdd1877d8d9515ab29f52c3db49ddd9612b5797bc071e6208def860c02e6f2"
         }
       ]
     },
@@ -415,7 +415,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "0d314eaeb69a761f68c4beb1ab7df0a2da1964538a8eb33c99cb9b819598d255"
+          "content": "97e9a7bc54f196b381c4ef71258cddcb2fb48f216df1ba015927c176bad57084"
         }
       ]
     },
@@ -639,6 +639,16 @@
         }
       ]
     },
+    {
+      "type": "file",
+      "name": "templates/.claude/skills/feature-adr/scripts/check-plan-completeness.mjs",
+      "hashes": [
+        {
+          "alg": "SHA-256",
+          "content": "41d85b012a9d6cc172eaa5d7777294f2c01fb54b752ed61c7cbe2d2cd7b79b67"
+        }
+      ]
+    },
     {
       "type": "file",
       "name": "templates/.claude/skills/frontend-design/LICENSE.txt",
@@ -795,7 +805,7 @@
       "hashes": [
         {
           "alg": "SHA-256",
-          "content": "8afa77bb421ea4052eec0feab5f49edc1f2b1cea69238971f8ef1b3c8328a483"
+          "content": "d9cda8df501188c894d083c5c44c240c9ebe9ba7e2d418302c448d73b089e051"
         }
       ]
     },

package/templates/.claude/skills/feature-adr/SKILL.md

@@ -100,6 +100,9 @@ ROUTER   →  REQUIREMENTS  →  RESEARCH  →   ADR         →  QCSD SWARM
                                            SPARC-GOAP PLAN
                                                 (all)
                                                    │
+                                        K2 PLAN-COMPLETENESS GATE
+                                       (all — exit 0 or Step 7 waits)
+                                                   │
                                                 Step 7
                                                  CODE
                                                 (all)
@@ -144,6 +147,26 @@ Step                         | S | M | L | XL | Model  | Agentic QE Skill
 10 Delivery Gate (OPT-IN)    | o | o | o | o  | cross-family of coder | 4 planes: regressions ‖ security ‖ code-quality ‖ product-honesty (o = runs only when explicitly requested; absent ⇒ byte-identical) |
 ```
 
+### K2 plan-completeness gate (Step-6/7 boundary, MANDATORY, all tiers)
+
+Step 7 does not start until the plan passes a SCRIPT — not a reviewer's impression:
+
+```bash
+node .claude/skills/feature-adr/scripts/check-plan-completeness.mjs features/<slug>
+```
+
+Pass the run's tier (`--tier=S|M|L|XL`): M/L/XL with no `03_adr/` FAILS rather than skipping C1/C2.
+Every tier — S included — writes a real `06_implementation_plan.md`; an inline-only S-tier checklist
+cannot be gated and is no longer permitted.
+
+`exit 0` → proceed to Step 7 · `exit 1` → return to Step 6 and fix every `FAIL C*` line · `exit 3`
+→ INCONCLUSIVE (inputs unreadable) — fix them and rerun. **Never proceed on a non-zero exit, and
+never read empty output as a pass**: the verdict is the last line
+(`K2 plan-completeness: PASS|FAIL|NOT-ESTABLISHED`), and its absence is not a verdict. The checkpoint
+banner's Gates line carries it: `🚦 Gates: K2 plan-completeness ✓ | ✗ | inconclusive`. Details and the
+per-check list: `modules/06-implementation-plan.md`. In the ultracode workflow the gate runs
+automatically and a non-PASS returns `phase: 'plan-gate-failed'` without dispatching the coder.
+
 ## DAG Dependencies
 
 Steps are NOT purely linear. The DAG defines what can run in parallel:
@@ -203,6 +226,14 @@ Execute protocol
 → Checkpoint N
 ```
 
+**Write discipline — applies to EVERY step that produces a document.** Open the step's artifact within
+your first ~12 tool calls as a SKELETON (its section headings, one line of intent each), then fill it one
+section per edit, no edit longer than ~120 lines. Never go more than 2 minutes without a tool call, and
+when you are unsure whether to read more or to write, WRITE. Reason: an executor silent for 180 seconds is
+killed by the runtime, and thinking time grows with accumulated history — so unbounded exploration before
+the first write is a deterministic death on a large repo, not bad luck. Each module restates this under
+its own `## Write discipline (the 180-second rule)` heading.
+
 ### 4. Final Verification
 After Step 8 (or Step 9 for L/XL) completes, verify:
 - All mandatory artifacts exist per tier
@@ -321,10 +352,17 @@ Two flags for integrating with the full agentic-qe package:
 
 | Flag | Mode | What it does |
 |------|------|-------------|
-| (none) | Reference | Condensed copies of 9 core skills, no install needed |
-| `--full-qe` | Direct | Full agentic-qe protocols for existing 9 core skills |
+| (none) | Reference | Condensed copies of 9 core skills, no install needed — the **no-install fallback** |
+| `--full-qe` | Direct | **RECOMMENDED DEFAULT when agentic-qe is installed.** Full agentic-qe protocols for the 9 core skills, and BOTH halves of the learning loop run: Step-0 recall = `memory_query("patterns/feature-adr/*")` **+** `dz recall`; Step-8 teach = the aqe `qe-outcome` store **+** `dz teach` |
 | `--full-qe-extended` | Direct Extended | Full protocols + 6 additional skills (chaos, security, performance, mutation, TDD, production-swarm) |
 
+**Which one to run:** if `aqe` is installed (`which aqe` or `node_modules/agentic-qe/`), run
+`--full-qe` — it is the recommended default at every tier, because the aqe pattern memory only
+contributes when the Direct half of the loop is active. Reference mode is the honest fallback for a
+machine without agentic-qe; it is not a lesser-quality choice there, it is the only correct one.
+This is a RECOMMENDATION about which flag to pass — the mode LOGIC is unchanged (flag + installed →
+direct; flag without install → WARN + reference).
+
 ### Installation
 
 ```bash
@@ -338,8 +376,8 @@ cd your-project && aqe init --auto
 ### Activation
 
 ```
-/feature-adr [описание фичи]                      # Reference Mode (default)
-/feature-adr --full-qe [описание фичи]             # Direct Mode: full protocols, same skills
+/feature-adr --full-qe [описание фичи]             # RECOMMENDED when agentic-qe is installed
+/feature-adr [описание фичи]                      # Reference Mode: the no-install fallback
 /feature-adr --full-qe-extended [описание фичи]     # Direct Extended: full protocols + extra skills
 ```
 
@@ -454,7 +492,8 @@ each step records the live panel via `dz statusline --fa-record`. This half NEVE
 
 The **agentic-qe MCP layer below is an ADDITIONAL enrichment, active only in Direct modes**
 (`--full-qe` / `--full-qe-extended`, needs `fleet_init`) — it adds semantic in-session recall on top of
-the always-on dz loop.
+the always-on dz loop. This is why `--full-qe` is the RECOMMENDED default wherever agentic-qe is
+installed: only then do BOTH recall halves run at Step 0 and BOTH teach halves at Step 8.
 
 When `{AGENTIC_QE_MODE}` = `direct` | `direct-extended`, the pipeline ALSO runs a recall → store cycle
 over agentic-qe's MCP pattern memory (`namespace: "learning"`, `fleet_init` first):
@@ -537,8 +576,9 @@ Everything from `--full-qe` plus 6 additional skills and up to 3 extra agents in
 
 | Scenario | Mode |
 |----------|------|
-| S/M tier features | Reference (always) |
-| L tier, standard QE | Reference or `--full-qe` |
+| agentic-qe installed (any tier) | `--full-qe` — the recommended default |
+| agentic-qe NOT installed | Reference (the fallback; the dz half of the loop still runs) |
+| S/M tier, agentic-qe installed | `--full-qe` |
 | L tier, high QE maturity | `--full-qe` |
 | XL tier features | `--full-qe` recommended |
 | XL + security-critical (banking, ФЗ-152) | `--full-qe-extended` recommended |
@@ -584,6 +624,7 @@ npx @dzhechkov/skills-feature-adr init --with-learning --knowledge-extractor
 | Ignore NO-GO verdict | Proceed despite Step 3.5 NO-GO | BLOCK — rework required |
 | Skip gap loop | No gap detection in Step 8 | Missing requirements coverage check |
 | Skip fleet QE for L/XL | No Step 9 for large features | BLOCK — fleet assessment mandatory |
+| Unbounded reading before the first write | The step's artifact still does not exist after ~12 tool calls | BLOCK — write the skeleton now; a silent executor is killed at 180 s |
 
 ## Checkpoint Format
 

package/templates/.claude/skills/feature-adr/modules/01-requirements.md

@@ -77,6 +77,32 @@ Explicitly state:
 - **Dependencies:** What must exist before this feature works
 - **Dependents:** What will break if this feature changes
 
+## Write discipline (the 180-second rule)
+
+An executor that returns from a tool call and then thinks in silence past **180 seconds** is killed by the
+runtime. Thinking time grows with the history already accumulated, so on a large repo "read everything,
+then write the document" is not a risk — it is a deterministic death, and nothing survives it, because
+nothing was ever on disk.
+
+MEASURED on this harness: the writing steps died **18 times out of 18** in the reading phase without ever
+writing a file. The control — same slice, same model, one added instruction to write a skeleton early —
+landed the skeleton 8 minutes in, on the first attempt, after six consecutive deaths.
+
+So, in this step:
+
+1. **Skeleton first — inside your first ~12 tool calls.** Write `features/<slug>/01_requirements.md` containing only the headings this step
+   requires (Stakeholders · Functional requirements · Non-functional requirements · Constraints ·
+   Scope boundaries · Open questions), one line of intent under each. (S-tier is inline and has no
+   file — this section does not apply to it.)
+2. **Then fill it one section per edit.** No single edit longer than ~120 lines. Every edit leaves the
+   file readable; none of them is allowed to wait for the section after it.
+3. **Never go more than 2 minutes without a tool call.** A thought that is getting long is the signal to
+   stop and write what you have — an edit is a checkpoint, not an interruption.
+4. **When you are unsure whether to read more or to write, WRITE.** A thin section refined later survives;
+   a perfect section you never reached does not.
+
+The skeleton is not a draft to apologise for. It is the artifact, opened early.
+
 ## Output
 
 ### For S-tier (inline)

package/templates/.claude/skills/feature-adr/modules/02-research.md

@@ -73,6 +73,32 @@ Alternative considered: [Y], rejected because [reasons].
 
 After both complete, synthesize results sequentially.
 
+## Write discipline (the 180-second rule)
+
+An executor that returns from a tool call and then thinks in silence past **180 seconds** is killed by the
+runtime. Thinking time grows with the history already accumulated, so on a large repo "read everything,
+then write the document" is not a risk — it is a deterministic death, and nothing survives it, because
+nothing was ever on disk.
+
+MEASURED on this harness: the writing steps died **18 times out of 18** in the reading phase without ever
+writing a file. The control — same slice, same model, one added instruction to write a skeleton early —
+landed the skeleton 8 minutes in, on the first attempt, after six consecutive deaths.
+
+So, in this step:
+
+1. **Skeleton first — inside your first ~12 tool calls.** Write `features/<slug>/02_research.md` containing only the headings this step
+   requires (Codebase patterns · External patterns · Library evaluation · Anti-patterns ·
+   Recommended approach), one line of intent under each — before the research agents report back,
+   not after.
+2. **Then fill it one section per edit.** No single edit longer than ~120 lines. Every edit leaves the
+   file readable; none of them is allowed to wait for the section after it.
+3. **Never go more than 2 minutes without a tool call.** A thought that is getting long is the signal to
+   stop and write what you have — an edit is a checkpoint, not an interruption.
+4. **When you are unsure whether to read more or to write, WRITE.** A thin section refined later survives;
+   a perfect section you never reached does not.
+
+The skeleton is not a draft to apologise for. It is the artifact, opened early.
+
 ## Output
 
 Create `features/<slug>/02_research.md` with:

package/templates/.claude/skills/feature-adr/modules/03-adr.md

@@ -167,6 +167,32 @@ Each ADR must trace back to specific requirements:
 | Placeholder text | Template hints or TODOs remain | Remove or fill before checkpoint |
 | Supersession in-place | Existing ADR rewritten | Mint a new ADR and link it |
 
+## Write discipline (the 180-second rule)
+
+An executor that returns from a tool call and then thinks in silence past **180 seconds** is killed by the
+runtime. Thinking time grows with the history already accumulated, so on a large repo "read everything,
+then write the document" is not a risk — it is a deterministic death, and nothing survives it, because
+nothing was ever on disk.
+
+MEASURED on this harness: the writing steps died **18 times out of 18** in the reading phase without ever
+writing a file. The control — same slice, same model, one added instruction to write a skeleton early —
+landed the skeleton 8 minutes in, on the first attempt, after six consecutive deaths.
+
+So, in this step:
+
+1. **Skeleton first — inside your first ~12 tool calls.** For each decision you have identified, write
+   `features/<slug>/03_adr/NNN-{decision-slug}.md` containing only the invariant-core headings
+   (Status · Context · Decision Drivers · Considered Options · Decision · Rationale · Consequences ·
+   Confirmation), one line of intent under each. A named-but-empty option beats an unwritten ADR.
+2. **Then fill it one section per edit.** No single edit longer than ~120 lines. Every edit leaves the
+   file readable; none of them is allowed to wait for the section after it.
+3. **Never go more than 2 minutes without a tool call.** A thought that is getting long is the signal to
+   stop and write what you have — an edit is a checkpoint, not an interruption.
+4. **When you are unsure whether to read more or to write, WRITE.** A thin section refined later survives;
+   a perfect section you never reached does not.
+
+The skeleton is not a draft to apologise for. It is the artifact, opened early.
+
 ## Output
 
 Create `features/<slug>/03_adr/` directory with:

package/templates/.claude/skills/feature-adr/modules/03.5-ideation-swarm.md

@@ -154,6 +154,31 @@ AM-N (source): <change>. Confirmation: <property> → test `test_name` (fails if
   an existence test and never fires in production.
 - A cheap lint holds the shape: every `AM-N` line must contain a `→ test ` token naming a test that exists.
 
+## Write discipline (the 180-second rule)
+
+An executor that returns from a tool call and then thinks in silence past **180 seconds** is killed by the
+runtime. Thinking time grows with the history already accumulated, so on a large repo "read everything,
+then write the document" is not a risk — it is a deterministic death, and nothing survives it, because
+nothing was ever on disk.
+
+MEASURED on this harness: the writing steps died **18 times out of 18** in the reading phase without ever
+writing a file. The control — same slice, same model, one added instruction to write a skeleton early —
+landed the skeleton 8 minutes in, on the first attempt, after six consecutive deaths.
+
+So, in this step:
+
+1. **Skeleton first — inside your first ~12 tool calls.** Write `features/<slug>/03.5_ideation_report.md` containing only the eight
+   section headings of §5 (Executive summary … Amendments), one line of intent under each — while
+   the swarm is still running, not after every agent has returned.
+2. **Then fill it one section per edit.** No single edit longer than ~120 lines. Every edit leaves the
+   file readable; none of them is allowed to wait for the section after it.
+3. **Never go more than 2 minutes without a tool call.** A thought that is getting long is the signal to
+   stop and write what you have — an edit is a checkpoint, not an interruption.
+4. **When you are unsure whether to read more or to write, WRITE.** A thin section refined later survives;
+   a perfect section you never reached does not.
+
+The skeleton is not a draft to apologise for. It is the artifact, opened early.
+
 ## Output
 
 Create `features/<slug>/03.5_ideation_report.md` with full QCSD report.

package/templates/.claude/skills/feature-adr/modules/04-ddd.md

@@ -92,6 +92,31 @@ Check that proposed domain model is compatible with existing code:
 - Can aggregates be implemented with existing ORM/framework?
 - Do relationships align with existing data access patterns?
 
+## Write discipline (the 180-second rule)
+
+An executor that returns from a tool call and then thinks in silence past **180 seconds** is killed by the
+runtime. Thinking time grows with the history already accumulated, so on a large repo "read everything,
+then write the document" is not a risk — it is a deterministic death, and nothing survives it, because
+nothing was ever on disk.
+
+MEASURED on this harness: the writing steps died **18 times out of 18** in the reading phase without ever
+writing a file. The control — same slice, same model, one added instruction to write a skeleton early —
+landed the skeleton 8 minutes in, on the first attempt, after six consecutive deaths.
+
+So, in this step:
+
+1. **Skeleton first — inside your first ~12 tool calls.** Write `features/<slug>/04_domain_model.md` containing only the headings this step
+   requires (Bounded context map · Ubiquitous language · Aggregates/Entities/VOs · Relationships ·
+   Codebase mapping), one line of intent under each.
+2. **Then fill it one section per edit.** No single edit longer than ~120 lines. Every edit leaves the
+   file readable; none of them is allowed to wait for the section after it.
+3. **Never go more than 2 minutes without a tool call.** A thought that is getting long is the signal to
+   stop and write what you have — an edit is a checkpoint, not an interruption.
+4. **When you are unsure whether to read more or to write, WRITE.** A thin section refined later survives;
+   a perfect section you never reached does not.
+
+The skeleton is not a draft to apologise for. It is the artifact, opened early.
+
 ## Output
 
 Create `features/<slug>/04_domain_model.md` with:

package/templates/.claude/skills/feature-adr/modules/05-architecture.md

@@ -118,6 +118,32 @@ GET /api/v1/{resource}/:id
   Errors: 401, 404
 ```
 
+## Write discipline (the 180-second rule)
+
+An executor that returns from a tool call and then thinks in silence past **180 seconds** is killed by the
+runtime. Thinking time grows with the history already accumulated, so on a large repo "read everything,
+then write the document" is not a risk — it is a deterministic death, and nothing survives it, because
+nothing was ever on disk.
+
+MEASURED on this harness: the writing steps died **18 times out of 18** in the reading phase without ever
+writing a file. The control — same slice, same model, one added instruction to write a skeleton early —
+landed the skeleton 8 minutes in, on the first attempt, after six consecutive deaths.
+
+So, in this step:
+
+1. **Skeleton first — inside your first ~12 tool calls.** Write `features/<slug>/05_architecture.md` containing only the headings this step
+   requires at your tier (C4 Context · C4 Container · C4 Component · Sequence flows · Data & storage ·
+   API design), one line of intent under each. An empty `mermaid` fence with a caption is a
+   heading; fill it later.
+2. **Then fill it one section per edit.** No single edit longer than ~120 lines. Every edit leaves the
+   file readable; none of them is allowed to wait for the section after it.
+3. **Never go more than 2 minutes without a tool call.** A thought that is getting long is the signal to
+   stop and write what you have — an edit is a checkpoint, not an interruption.
+4. **When you are unsure whether to read more or to write, WRITE.** A thin section refined later survives;
+   a perfect section you never reached does not.
+
+The skeleton is not a draft to apologise for. It is the artifact, opened early.
+
 ## Output
 
 ### M-tier