@tacuchi/agent-workflow-cli 20.23.0 → 20.24.0

package/skills/w/loops/CHASSIS.md

@@ -6,27 +6,27 @@ This document is the **common engine** of the Workline loops: the doctrine every
 
 The **5 loops** run this engine; each adds only its deltas:
 
-- [`spec-refine-loop`](spec-refine-loop/LOOP.md) — refines the **spec** in place; deltas: spec gap taxonomy, analyze gate, `## UI spec` via the `ui-design` capability.
-- [`plan-new-loop`](plan-new-loop/LOOP.md) — generates the **plan** from the spec; deltas: rich plan + plan gap taxonomy (+ per-screen design SPECs when the plan includes UI).
-- [`plan-refine-loop`](plan-refine-loop/LOOP.md) — refines the **plan** in place (auxiliary, not mandatory); reuses the gap taxonomy + coherence gate of `plan-new-loop`. It is to `plan-new` what `spec-refine` is to `spec-new`.
-- [`plan-exec-loop`](plan-exec-loop/LOOP.md) — **executes** the plan: code/DB/git, a single session per run, per-phase progress in the plan-doc, no auto-export. Applies the policies in [`CODE-POLICIES.md`](CODE-POLICIES.md).
-- [`quick-loop`](quick-loop/LOOP.md) — the engine with **minimal ceremony** (the prompt *is* the objective); also applies [`CODE-POLICIES.md`](CODE-POLICIES.md) (proportional gate).
+- [`spec-refine-loop`](spec-refine-loop/LOOP.md) — refines the **spec** in place.
+- [`plan-new-loop`](plan-new-loop/LOOP.md) — generates the **plan** from the spec.
+- [`plan-refine-loop`](plan-refine-loop/LOOP.md) — refines the **plan** in place (auxiliary).
+- [`plan-exec-loop`](plan-exec-loop/LOOP.md) — **executes** the plan: code, DB and git.
+- [`quick-loop`](quick-loop/LOOP.md) — **minimal ceremony** (the prompt *is* the objective).
 
-## Persistent objective
+The two that edit code also apply [`CODE-POLICIES.md`](CODE-POLICIES.md).
 
-A loop **is a persistent objective**: it exists to fulfill the `SESSION.Objective` declared at start, and **it is not finished until the convergence gate confirms the objective was met**. Gap-driven iteration is the *method*; the artifacts are the *record*; the persistent objective is the *frame* that governs them.
+## Persistent objective
 
-This is **harness-agnostic doctrine**, not a host dependency: "don't stop until convergence" is sustained by the loop itself (its `repeat:` + the convergence gate), not by a host hook — and it **leaves a durable record** (artifact-first) that survives compaction and resume. *(Rationale and the `/goal` analogy: see the design, `workflow-loops/chassis.md`.)*
+A loop **is a persistent objective**: it exists to fulfill the `SESSION.Objective` declared at start, and **it is not finished until the convergence gate confirms the objective was met**. Gap-driven iteration is the *method*; the artifacts are the *record*; the objective is the *frame*.
 
-> Each heir instantiates the frame: `spec-refine` pursues the spec; `plan-new`/`plan-refine` pursue the plan up to their gate; `plan-exec` pursues the plan up to its final validation; `quick-loop` is the most direct embodiment (the prompt *is* the objective).
+"Don't stop until convergence" is sustained by the loop itself — its `repeat:` plus the convergence gate — never by a host hook, which is what makes it harness-agnostic. Each heir instantiates the frame: `spec-refine` pursues the spec; the plan loops pursue the plan up to their gate; `plan-exec` up to its final validation; `quick-loop` most directly of all (the prompt *is* the objective).
 
-> **Inter-turn continuity.** The same `CHECKPOINT`+resume also governs the **next prompt**: the objective persists **across turns**, not only within a run. The canonical rules (command = new work line · re-run = `create_or_resume` · bare prompt = continue the most recent session · reopening closed sessions · consented escalation) live in [`../SKILL.md`](../SKILL.md) § *Operating context* — **single source**; this engine executes them via *Compact / resume* (case 3).
+> **Inter-turn continuity.** The same `CHECKPOINT`+resume governs the **next prompt**: the objective persists **across turns**, not only within a run. The canonical rules (command = new work line · re-run = `create_or_resume` · bare prompt = continue the most recent session · reopening closed sessions · consented escalation) live in [`../SKILL.md`](../SKILL.md) § *Operating context* — **single source**; this engine executes them via *Compact / resume* (case 3).
 
 ## Verification-first
 
 The persistent objective needs a **checkable done-condition** — otherwise the loop cannot know when it is done (or chases a target it invented). That condition is **seeded BEFORE executing**, never improvised at the end: it is **generalized TDD**. Together with artifact-first (next section) these are the **two seeds** of every gap/phase: *how will I know it worked* + *what am I about to do*.
 
-**Where it lives:** in `SESSION.Success criteria` (see [`../artifacts/artifacts-core/SESSION.md`](../artifacts/artifacts-core/SESSION.md)) — a `[ ]` checklist of **falsifiable** criteria (that *can* fail). `CHECKPOINT.Pending/Completed` tracks the **red→green** progress. Two forms, by deliverable:
+**Where it lives:** in `SESSION.Success criteria` — a `[ ]` checklist of **falsifiable** criteria (that *can* fail). `CHECKPOINT.Pending/Completed` tracks the **red→green** progress. Two forms, by deliverable:
 
 | Deliverable | Criterion = | Cycle |
 |---|---|---|
@@ -35,18 +35,13 @@ The persistent objective needs a **checkable done-condition** — otherwise the
 | spec / plan | **rubric** = the document's acceptance criteria (referenced, not duplicated) | rubric |
 | analysis / design | **rubric falsifiable by inspection** (e.g. "every affected site with `file:line`"; "each decision: rationale + ≥1 alternative") | rubric |
 
-- **Form and weight scale** (quick's minimal ceremony preserved): a chore = "existing tests/build stay green" (one line); a feature = real acceptance tests. The rule is "**always declare the check before**", not "always write new tests".
-- **Subjective deliverable** (analysis/design): the AI **proposes** the rubric and the **human ratifies** it (structured-choice) before pursuing it.
-- **Unresolvable criterion** (no evidence, DB unavailable): closes as `inconclusive` and the loop **degrades** (asks the human, or defers to `Open questions`/`BACKLOG`) — **never iterates against a fake target**.
-
-> The **convergence gate** (section *Convergence / exit*) is, operationally, **"all `Success criteria` green"**. The per-heir gates (analyze gate; plan coherence — plan-new and plan-refine; final validation; proportional spot validation) are **instances** of it, with the criteria seeded at start.
+- **Form and weight scale**: a chore = "existing tests/build stay green" (one line); a feature = real acceptance tests. The rule is "**always declare the check before**", not "always write new tests".
+- **Subjective deliverable** (analysis/design): the AI **proposes** the rubric and the **human ratifies** it before pursuing it.
+- **Unresolvable criterion** (no evidence, DB unavailable): closes as `inconclusive` and the loop **degrades** — **never iterates against a fake target**.
 
-**Gate integrity (anti-gaming + independent verification).** The gate only counts if it is not gamed to pass. The loop does **not**:
+> The **convergence gate** is, operationally, **"all `Success criteria` green"**. The per-heir gates are **instances** of it, with the criteria seeded at start.
 
-- modify the check or loosen a `Success criterion` to force green;
-- weaken, delete or skip tests/validations;
-- use trivial or tautological asserts that always pass (the expected value comes from an independent source, never from the output itself);
-- patch the test instead of fixing the cause (prefer fixing production code).
+**Gate integrity (anti-gaming + independent verification).** The gate only counts if it is not gamed to pass. The loop does **not**: modify the check or loosen a `Success criterion` to force green; weaken, delete or skip tests/validations; use trivial or tautological asserts that always pass (the expected value comes from an independent source, never from the output itself); patch the test instead of fixing the cause (prefer fixing production code).
 
 Facing a real blocker it **stops and reports it** (→ `Open questions`/`BACKLOG`) instead of gaming the metric. The verdict counts **only the check's output, never the implementer's self-declaration**: when the deliverable warrants it, the final verification is an **independent** pass (subagent or clean re-read) that does not assume the implementation is correct — *only command output counts*.
 
@@ -66,9 +61,9 @@ The loop works **artifact-first**: the artifact is **seeded before** executing a
 2. **EXECUTE.** Resolve the gap / run the phase / edit the code.
 3. **AFTER — bring to actual state.** `CHECKPOINT.Pending → Completed`; `DECISION` records the non-obvious **as it is decided**; `BACKLOG` **only if** something is deferred/follow-up (`session-close` no longer fabricates an empty BACKLOG).
 
-> The artifact expresses the **intent** (Pending/Next, before) and then the **result** (Completed/DECISION, after), at **every** gap/phase boundary — not only on `Compactar`/`Cerrar`. Session artifacts are the run's live log; the spec/plan is the **guiding base**.
+> The artifact expresses the **intent** (before) and then the **result** (after), at **every** gap/phase boundary — not only on `Compactar`/`Cerrar`. Session artifacts are the run's live log; the spec/plan is the **guiding base**.
 
-> **Fixed form (hard rule):** an artifact keeps its template's `##` headings **exactly** and is updated **in place** — appending a **duplicate heading** is a contract violation. When a scaffolded section is filled, its `<!-- … -->` guidance comment is **replaced** by the real content. Canonical headings per artifact: its template under [`../artifacts/`](../artifacts/) (CHECKPOINT contract: `Completed` · `Pending / Next`; `Open questions` only while live doubts exist).
+> **Fixed form (hard rule):** an artifact keeps its template's `##` headings **exactly** and is updated **in place** — appending a **duplicate heading** is a contract violation. A filled section **replaces** its `<!-- … -->` guidance comment. Canonical headings per artifact: its template under [`../artifacts/`](../artifacts/) (CHECKPOINT: `Completed` · `Pending / Next`; `Open questions` only while live doubts exist).
 
 ## Gap-driven convergent engine
 
@@ -82,126 +77,80 @@ The common cycle — each heir instantiates it in its `## Sequence` with its own
 
 ## Internal sessions (managed) — one session per run
 
-The loop creates and manages its session under `.workflow/sessions/`. **The user never creates it.** **A single session per run**, owning the run: it keeps progress live (`CHECKPOINT`) and enables resume. Artifacts: `SESSION.md` · `CHECKPOINT.md` (· `BACKLOG.md` only if something is deferred; code-editing loops add `DECISION` and `SCRIPTS.sql`). Each heir declares its descriptor and `Type` in its own `## Internal sessions`.
-
-> **INLINE research** — investigation is **not** a separate session: it is an activity **inside the current session** that writes its artifacts (`ANALYSIS-FILE`/`CONCLUSIONS`, + read-only `SCRIPTS.sql` if it queries DB) **into the run's own session folder**. See *Research: autonomy, scope & failure*.
-
-> The flow's input document (spec/plan) **never** goes inside a session; it lives in `docs/`.
+The loop creates and manages its session under `.workflow/sessions/`; **the user never creates it**. **A single session per run**: it keeps progress live (`CHECKPOINT`) and enables resume. Artifacts: `SESSION.md` · `CHECKPOINT.md` (· `BACKLOG.md` only if something is deferred; code-editing loops add `DECISION` and `SCRIPTS.sql`). Each heir declares its descriptor and `Type` in its own `## Internal sessions`.
 
-### Session numbering (hard rule)
+> Research is **inline** — an activity inside this same session, never a session of its own — and the flow's input document (spec/plan) **never** goes inside a session; it lives in `docs/`.
 
-The **CLI owns the number**: `aw session-create` prepends a **global, sequential** `NNN` by scanning **all** sessions under `.workflow/sessions/` (any type). The caller passes **only the descriptor** via `--name` — **never** a number. Numbering neither restarts per type nor collides, and every folder is **self-describing**: `NNN-<slug>-<flow>` (e.g. `002-correo-otp-spec-refine`, `003-correo-otp-plan-new`, `004-correo-otp-plan-exec`, `005-validacion-correo-quick`).
+**CLI**: `aw session-create --type <type> --name <slug>-<flow> --objetivo "<one-line objective>"` opens it · `aw checkpoint-write` / `aw checkpoint-read` keep it resumable · `aw session-close` closes it and upserts its HISTORY.md row.
 
-> `<run>` = the session's **descriptor** (no number), always shaped **`<slug>-<flow>`**: `<slug>-spec-refine`, `<slug>-plan-new`, `<slug>-plan-refine`, `<slug>-plan-exec`, `<slug>-quick`. The `<slug>` is **descriptive** and comes from the flow's input doc — `docs/specs/NNN-spec-<slug>.md` for spec-refine/plan-new; `docs/plans/PPP-plan-<slug>.md` for plan-refine/plan-exec; the prompt for quick — so the folder says at a glance what it is about, not just which flow created it. Research being **inline** in this same session, there are no child `*-research-*` sessions to number (compat: old ones are historical).
->
-> **Resume**: locate the existing session by **scanning** `.workflow/sessions/` for descriptor + `## Origin` (which spec/plan), **not** by reconstructing the number (global, not derivable from the artifact). `aw session-resume --code <NNN | folder>` resolves both forms.
-
-**CLI**:
-
-- `aw session-create --type <type> --name <slug>-<flow> --objetivo "<one-line objective>"` → creates `NNN-<slug>-<flow>` / `aw session-resume --code <…>` (detects `CHECKPOINT`).
-- `aw checkpoint-write` / `aw checkpoint-read` for resume.
-- `aw session-close` on close — also upserts the session's HISTORY.md row (the durable record; sessions/ is gitignored). Non-fatal: on `history_error` in its output, re-run `aw history-update --code <NNN> --state closed`. `aw session-artifacts` to inspect.
-- **Reopen to continue** (operating context, row 2): `aw session-resume --code <NNN> --reopen` reactivates a **closed** session (removes `.closed` → active) to keep working in it; without `--reopen`, resume is read-only. To detect the most recent closed one: `aw resume-summary --include-recent-closed` (or `aw sessions --state all`).
+> The caller passes **only the descriptor** via `--name` — **never** a number; the CLI owns the global `NNN`. How it is assigned, how a session is located or reopened, and how a failed history upsert is repaired, are in the `sessions` module.
 
 ## Ask-vs-research rule (the discriminator)
 
 For every gap, a single question picks the resolver:
 
-> *"Was this already established in the current conversation?"* → **adopt it** (see *Adopted context*) — never re-ask or re-research settled conclusions.
+> *"Was this already established in the current conversation?"* → **adopt it** (`adopted` module) — never re-ask or re-research settled conclusions.
 > *"Can I answer this by reading the repo/data?"* → **research** (autonomous).
-> *"Can I only answer it by RUNNING a small experiment?"* → **probe** (see *Proof of concept*).
+> *"Can I only answer it by RUNNING a small experiment?"* → **probe** (`probe` module).
 > *"Does it depend on what the user wants?"* → **ask the human** (structured-choice).
 
-## Adopted context (host as producer)
-
-The host is not only the loop's executor — it is a legitimate **producer** of input. Conclusions **already established in the current conversation** (a host-native analysis, answers the user already gave, a plan built with the host's planner) count as **completed research**: on entry, **adopt** them — seed `SESSION.Objective`/`Success criteria`, reference them in `CONCLUSIONS`, record provenance (`## Origin` = adopted from the host conversation) — never re-derive or re-ask them.
-
-- Adoption is **transcription, not trust**: the convergence gate still verifies adopted conclusions (*gate integrity*); anti-duplicate still applies.
-- Materialization pattern = the quick escalation's (single-pass, **NO RESEARCH**), inverted: **host → flow**.
-- Gap signals already resolved by adopted context do **not** fire (e.g. "ambiguous requirements" after a host pre-analysis).
-- To persist finished work without a loop: [`/w:persist`](../commands/persist.md) (transversal).
-
 ## Research: autonomy, scope & failure
 
-Investigation is **inline**: an activity **inside the run's current session**, never a separate session. It writes its artifacts (`ANALYSIS-FILE` → `CONCLUSIONS`, + read-only `SCRIPTS.sql` if it queries DB) into the **session's own folder**.
-
-- **Autonomous**: the AI investigates inline and reports **without asking permission**. The human learns of it at integration time (in the flow's decision record — each heir names its own) and keeps control via the `flow` control.
-- **Scope**: the current conversation (*adopted context* — settled conclusions are reused, never re-derived) + workspace + associated repos (sources) + DB MCPs.
-- **DB rule** (the single exception to autonomy):
-  1. **MCP choice**: if the gap needs DB and there is **>1 candidate MCP with no configured default**, the AI asks which one to use. That question goes through the **same structured-choice** as a **content question** (counts inside the ≤3 + `flow` limit), **before** running queries. A single MCP or a default → no question.
-  2. Write the queries **first** into the session's `SCRIPTS.sql`.
-  3. Execute them **read-only** via MCP (respect `sql-mutation-guard`: never DML/DDL).
-- **Inconclusive research** (DB unavailable, insufficient evidence, unresolvable factual gap):
-  - The investigation closes with status **`inconclusive`** in `CONCLUSIONS` and reports why.
-  - The loop **degrades** the gap: to a **human question** (next batch → the flow's decision record) or, failing that, **defers** it to the flow doc's `## Open questions` (spec/plan) — or the session's `BACKLOG` when the flow has no doc (quick).
-  - The gap is marked **"already tried via research"** (`attempts[gap]++`, `MAX` cap) so `detect_gaps` does **not** re-fire it in a loop → guarantees convergence.
-
-## Proof of concept (probe)
+Investigation is **inline**: an activity **inside the run's current session**, never a separate session. It writes its artifacts (`ANALYSIS-FILE` → `CONCLUSIONS`, + read-only `SCRIPTS.sql` if it queries DB) **into the session's own folder**.
 
-A **probe** (PoC / spike) is the resolver for **executable doubt**: research *reads*, a probe *runs* — an **atomic, throwaway-by-default** experiment answering **one falsifiable question** (does this connection / SDK / UI behavior work as assumed?). De-risk atomic parts **early**, never everything at the end.
-
-- **When**: risky assumption + not answerable by reading + failure would invalidate downstream work. Proposed via **structured-choice**.
-- **Lifecycle** (verification-first applies to the probe): seed question + pass/fail check **BEFORE** → run minimal → verdict in `CONCLUSIONS` (consequences → `DECISION`) → **discard** or promote to a real task/test.
-- **Isolation**: probe code lives in the **session folder** (gitignored) — never the source tree, **never committed**; DB probes are read-only (never DDL/DML).
-- A **failed probe is a finding, not a failure** — report it; the human decides if the plan reshapes.
+- **Autonomous**: the AI investigates inline and reports **without asking permission**. The human learns of it at integration time and keeps control via the `flow` control.
+- **Scope**: the current conversation (settled conclusions are reused, never re-derived) + workspace + associated repos + DB MCPs.
+- **DB rule** — the single exception to autonomy: it lives in the `db` module and is loaded **before** any query runs.
+- **Inconclusive research** (DB unavailable, insufficient evidence, unresolvable factual gap): the investigation closes with status **`inconclusive`** in `CONCLUSIONS` and reports why. The loop **degrades** the gap — to a human question, or failing that to the flow doc's `## Open questions` (the session's `BACKLOG` when the flow has no doc) — and marks it **"already tried via research"** (`attempts[gap]++`, `MAX` cap) so `detect_gaps` does **not** re-fire it in a loop. That is what guarantees convergence.
 
 ## Structured-choice (design & batching)
 
 **Canonical rule (single source — the rest of the corpus only references it):** *structured-choice* = **≤3 content questions + 1 `flow` control**, always. Per-harness binding in [`../harness/HARNESS.md`](../harness/HARNESS.md) (Claude Code: `AskUserQuestion`, max 4 questions/call; without structured choice it degrades to **numbered markdown**).
 
-- Since the `flow` control is **always** present → **≤3 content questions + 1 `flow` control**.
-- **`flow` control** (lifecycle, always present): `Compactar` | `Cerrar`. Answering only the content questions (not touching `flow`) = keep iterating.
-- **Proactive raise**: under context pressure the loop does not wait for the next content batch — it raises the structured-choice itself, with `Compactar` as the recommended action (see *Compact / resume* → *Self-regulation (proactive compaction)*).
-- **Content questions** can be:
-  - human doubts (non-factual gaps);
-  - MCP choice (DB rule) — before running queries;
-  - at **convergence**, the loop's own closing action — **each heir defines it in its *Convergence / exit*** (e.g. `Guardar especificación refinada` · `Cerrar tarea`) — | `Preguntar algo más`.
-- **Batching**: group up to 3 human gaps in one call. With more than 3 pending, prioritize (the ones that unblock other gaps first) and defer the rest to the next round.
-- **Recommended answer per question**: every content question **always** carries the AI's recommended answer — as the first option (marked *recommended*) in `AskUserQuestion`, or flagged in the numbered-markdown fallback. Never ask "cold": the human ratifies or corrects a proposal, never starts from zero. The AI recommends based on what it researched (ask-vs-research rule), never on an empty default.
+- **`flow` control** (lifecycle, always present): `Compactar` | `Cerrar`. Answering only the content questions = keep iterating. Under context pressure the loop **raises the choice itself**, with `Compactar` recommended.
+- **Content questions** are: human doubts (non-factual gaps) · MCP choice (DB rule), before running queries · at **convergence**, the loop's own closing action — each heir defines it in its *Convergence / exit* (e.g. `Guardar especificación refinada` · `Cerrar tarea`) — | `Preguntar algo más`.
+- **Batching**: up to 3 human gaps per call; with more pending, prioritize the ones that unblock others and defer the rest.
+- **Recommended answer per question**: every content question **always** carries the AI's recommended answer — the first option, marked *recommended*. Never ask "cold": the human ratifies or corrects a proposal, never starts from zero. The recommendation comes from what was researched (ask-vs-research rule), never from an empty default.
 
-> **Label language:** the literal option labels (`Compactar`, `Cerrar`, `Guardar plan`, …) are **canonical product strings** — present them **verbatim**; they are user-facing, authored in the product's user language (Spanish). All other user-facing output follows [`../SKILL.md`](../SKILL.md) § *Language policy*.
+> **Label language:** the literal option labels (`Compactar`, `Cerrar`, `Guardar plan`, …) are **canonical product strings** — present them **verbatim**. All other user-facing output follows [`../SKILL.md`](../SKILL.md) § *Language policy*.
 
 ## Compact / resume
 
 Resume **keys off the `CHECKPOINT`** of the run's session, not the existence of a separate file. Three cases when the flow's command runs over an input:
 
 1. **In progress** (a `CHECKPOINT.md` exists in the session) → resume from the recorded progress (resolved gaps, Q&A, `attempts`, in-flight inline research).
-2. **No progress** (no CHECKPOINT and the input doc does **not** have the flow's prior-work mark) → start from zero reading the input doc (plus any *adopted context* — settled in-conversation conclusions are input, not something to re-derive).
-3. **Already converged / re-run on demand** (no open CHECKPOINT but the doc **already has** the mark) → **first-class operation**: while the flow stays in its stage, re-running the command over the same input **as many times as needed** is supported. `create_or_resume` finds the existing session — typically **closed** after convergence — by descriptor + `## Origin` and **reopens** it (see *Internal sessions*: detection via `aw sessions --state all` / `aw resume-summary --include-recent-closed`, reopening via `aw session-resume --code <NNN> --reopen`); incremental work reading the **doc itself**.
-
-> Each heir defines its **prior-work mark** and its own legacy tolerances: spec-refine keys off the spec's frontmatter `status`; plan-refine off `## Refinement decisions` in the plan; plan-exec off the plan-doc's `- [x]` checkboxes; quick has no doc (resume by CHECKPOINT only).
+2. **No progress** (no CHECKPOINT and the input doc does **not** have the flow's prior-work mark) → start from zero reading the input doc (plus any settled in-conversation conclusions, which are input, not something to re-derive).
+3. **Already converged / re-run on demand** (no open CHECKPOINT but the doc **already has** the mark) → **first-class operation**: while the flow stays in its stage, re-running the command over the same input **as many times as needed** is supported. `create_or_resume` finds the existing session — typically **closed** after convergence — by descriptor + `## Origin` and **reopens** it; incremental work reading the **doc itself**.
 
-> **`Compactar`** (the `flow` control, across all 3 cases) → write `CHECKPOINT.md` in the session (in-flight progress, remaining gaps, Q&A, `attempts`) → trigger the harness **compaction** (Claude Code: `/compact`; see [`../harness/HARNESS.md`](../harness/HARNESS.md)) → resume by reading the checkpoint.
+> Each heir defines its **prior-work mark**: spec-refine the spec's frontmatter `status`; plan-refine `## Refinement decisions`; plan-exec the plan-doc's `- [x]` checkboxes; quick has no doc (CHECKPOINT only).
 
-### Self-regulation (proactive compaction)
-
-`Compactar` is not only reactive: the loop **watches its own context pressure** and raises compaction itself.
-
-- **Signal**: the host's context-pressure signal when it exists (see the *compaction* capability in [`../harness/HARNESS.md`](../harness/HARNESS.md)); with no signal the fallback is **qualitative** — at batch/phase boundaries of an already-long run, ask *"would a fresh reader need the CHECKPOINT to continue?"*. Doctrine fixes **no numeric thresholds** (harness-agnostic).
-- **Modes** — config `[compaction]` table in `.workflow/skills.toml`, key `mode` (values `confirm` | `auto`):
-  - **`confirm`** (default, also with no config): raise a **proactive structured-choice** whose `flow` control carries `Compactar` as the recommended action — the human ratifies; consent is never skipped.
-  - **`auto`** (opt-in): write `CHECKPOINT.md`, then trigger the host's compaction binding **without asking**. Viable only where the host has a **non-interactive** mechanism (see the *Harness binding matrix*); otherwise it **degrades to `confirm`**.
-- **Invariant — CHECKPOINT before compacting**: in every mode, `CHECKPOINT.md` is written (or verified fresh) **before** any compaction fires; resume keys off it (the 3 cases above).
+> **`Compactar`** (the `flow` control, across all 3 cases) → write `CHECKPOINT.md` in the session → trigger the harness compaction → resume by reading the checkpoint. **CHECKPOINT before compacting** is invariant. The proactive raise, its two modes and their host binding live in the `compaction` module.
 
 ## Convergence / exit
 
-- **No material gaps** → **convergence gate** (read-only) = **`Success criteria` green** (*verification-first*). Whatever fails **comes back as a gap**; if it passes → the loop **flips the green criteria** in `SESSION.md` (`- [ ]` → `- [x]`) and offers its closing action. The checklist must reflect the real final state: a criterion left unchecked at `finalize` needs an explicit reason (`Open questions`/`BACKLOG`). The heirs are **instances** of the same gate: `spec-refine` = analyze gate, `plan-new` and `plan-refine` = plan coherence, `plan-exec` = final validation, `quick` = proportional spot validation.
-- `Cerrar` (the `flow` control, at any time) → `finalize`. **`finalize` always persists `CHECKPOINT.md`** (resumable) and, **only if something was deferred/follow-up**, writes `BACKLOG.md` (close reason + the deferred items); closes the session and reports. Progress survives even without a prior `Compactar`.
+- **No material gaps** → **convergence gate** (read-only) = **`Success criteria` green** (*verification-first*). Whatever fails **comes back as a gap**; if it passes → the loop **flips the green criteria** in `SESSION.md` (`- [ ]` → `- [x]`) and offers its closing action. The checklist must reflect the real final state: a criterion left unchecked at `finalize` needs an explicit reason. Each heir names its own instance of this gate.
+- `Cerrar` (the `flow` control, at any time) → `finalize`. **`finalize` always persists `CHECKPOINT.md`** (resumable) and, **only if** something was deferred, writes `BACKLOG.md` (close reason + the deferred items); closes the session and reports. Progress survives even without a prior `Compactar`.
 
 ## docs/ boundary — no auto-export (hard rule)
 
-A loop writes into `docs/` **only** its own flow's doc (spec-refine: `docs/specs` · plan-new/plan-refine/plan-exec: `docs/plans` · quick: **none** — it never touches `docs/`). No loop **graduates/promotes artifacts** into `docs/`: everything else (migrations → `docs/scripts`, manuals → `docs/manuals`, diagrams → `docs/diagrams`, …) is done by the separate **`export-*`** skills, as an explicit later step. Artifacts stay in their sessions until then. If a task creates a tool/utility, the ambient skill `creating-tools` documents it in `docs/tools` (auto-discovered by its `description`; Workline is **indifferent** — it does not bind it).
+A loop writes into `docs/` **only** its own flow's doc (spec-refine: `docs/specs` · the three plan flows: `docs/plans` · quick: **none** — it never touches `docs/`). No loop **graduates/promotes artifacts** into `docs/`: migrations → `docs/scripts`, manuals → `docs/manuals`, diagrams → `docs/diagrams` are done by the separate **`export-*`** skills, as an explicit later step; artifacts stay in their sessions until then. A task that creates a tool/utility has it documented in `docs/tools` by the ambient `creating-tools` skill (auto-discovered; Workline does not bind it).
 
-## Code-editing loop policies → CODE-POLICIES.md
+## Conditional modules
 
-The loops that **edit code** (`plan-exec-loop`, `quick-loop`) additionally run the policies in [`CODE-POLICIES.md`](CODE-POLICIES.md) — **safe git** (verified branch + proposed commits) · **DB scripts-only** · **closing review gate** (proportional in quick). They order it read from their `## Inherits` **together with this chassis**; the document loops (spec-refine, plan-new, plan-refine) do **not** load it — that is why it lives in a separate doc.
+The engine above is what every run needs. The branches below apply only under their signal, and `aw context-plan` returns them exactly then — never just in case:
+
+- `adopted` → [`../modules/ADOPTED-CONTEXT.md`](../modules/ADOPTED-CONTEXT.md) · `probe` → [`../modules/PROBE.md`](../modules/PROBE.md) · `db` → [`../modules/DB-RESEARCH-RULE.md`](../modules/DB-RESEARCH-RULE.md)
+- `compaction` → [`../modules/COMPACTION.md`](../modules/COMPACTION.md) · `sessions` → [`../modules/SESSION-NUMBERING.md`](../modules/SESSION-NUMBERING.md)
+- `code` → [`CODE-POLICIES.md`](CODE-POLICIES.md), which the two code-editing loops order read from their `## Inherits`
 
 ## Reference resolution (global layout rule) — and what the chassis is NOT
 
 Applies to **every** relative reference in the doctrine — never repeated per link:
 
-1. **Normal install** (the `w/` tree): the relative path resolves as-is (`../CHASSIS.md`, `../../commands/spec-new.md`).
-2. **Synthesized command skills** (hosts without a commands dir — Codex/Warp/Oz): each command installs as a **sibling** skill `w-<command>/` with its references rewritten into the bundle (`../loops/…` → `../w/loops/…`); the `w/` tree stays intact, so loop-to-loop references resolve as a normal install.
-3. A reference that does not resolve = **optional deep-dive** — this engine's doctrine is self-contained.
+1. **Normal install** (the `w/` tree): the relative path resolves as-is.
+2. **Synthesized command skills** (hosts without a commands dir): each command installs as a sibling skill `w-<command>/` with its references rewritten into the bundle (`../loops/…` → `../w/loops/…`).
+3. A reference that does not resolve = **optional deep-dive** — this engine is self-contained.
+
+> `aw context-plan` hands back **absolute** paths, so a run that asks the CLI for its read-set resolves no relative reference at all. This rule is the fallback for one that does not ask.
 
-The chassis **is not a skill** (no frontmatter; never invoked nor bound via `.workflow/skills.toml`): it enters the context only because a loop orders it read from its `## Inherits`. It does not define flow, deliverable or gap taxonomy — that belongs to each heir.
+The chassis **is not a skill** (no frontmatter; never invoked nor bound): it enters the context only because a loop orders it read from its `## Inherits`. It does not define flow, deliverable or gap taxonomy — that belongs to each heir.

package/skills/w/loops/CODE-POLICIES.md

@@ -9,15 +9,6 @@ They apply to **`plan-exec-loop`** (per plan phase) and **`quick-loop`** (the si
 - **Rejected commit**: the changes **stay in the working tree** (never reverted). Re-proposing / editing the message is allowed. Record in `CHECKPOINT` + `BACKLOG` that the phase/task remained **uncommitted** (resumable).
 - **Between-phase precondition** (plan-exec): `branch-check` validates branch *identity*, **not** working-tree *cleanliness*. Before starting the next phase, each source's working tree must be **clean** (committed) or explicitly **acknowledged** as "uncommitted changes from phase N" — so two phases never co-mingle in one commit.
 
-## DB scripts-only — the AI never executes DML/DDL
-
-Distinguished by **execution**, not by file (see the [`SCRIPTS.sql`](../artifacts/artifacts-core/SCRIPTS.sql) schema):
-
-- **Read-only queries** (diagnosis/validation) → `SCRIPTS.sql` (session artifact); the AI **does** execute them read-only via MCP (`sql-mutation-guard`).
-- **DDL/DML migrations** (schema/data changes) → the AI **drafts them in `SCRIPTS.sql`** (session artifact) but **NEVER executes them**.
-
-> Mutating SQL **stays in the session**; it is never moved to `docs/`. Its promotion to `docs/scripts/` (forward + rollback) is done by a separate `export-*`, never by the loop.
-
 ## Closing review gate (conventions, pre-commit)
 
 After validation (of the phase in plan-exec; of the task in quick, proportional) and **before proposing its commits** (also on an early `Cerrar`, before proposing the pending commits), the diff passes a **closing review gate**:
@@ -36,3 +27,7 @@ Only with the gate green are the commits proposed.
 ## Location
 
 Same as the chassis: code-editing loops reference it as `../CODE-POLICIES.md` — the `w/loops/` tree is installed intact on every host (chassis § *Reference resolution*).
+
+## Conditional modules
+
+- `db` — the DB scripts-only rule → `../modules/DB-SCRIPTS-ONLY.md`

package/skills/w/loops/plan-exec-loop/LOOP.md

@@ -110,10 +110,6 @@ On either return path: `CHECKPOINT` records the state reached and the trigger, t
 
 Full policy in [`../CODE-POLICIES.md`](../CODE-POLICIES.md) (§ *Safe git*: branch-check before editing, rejected commit — changes stay + get recorded —, working-tree precondition between phases). **Inline:** before editing, verify each source's expected branch (`aw check-branch --source <alias>`; on mismatch → pause and resolve with the human); at each phase close and **after the review gate** (Delta 5), **proposed commits per source** (approve first) — never `push`/`--amend`/`--no-verify`.
 
-## Delta 3 — DB policy: **the AI never executes DML**
-
-Full policy in [`../CODE-POLICIES.md`](../CODE-POLICIES.md) (§ *DB scripts-only*). **Inline:** read-only queries → the session's `SCRIPTS.sql`, executed via MCP (`sql-mutation-guard`); DDL/DML migrations → the AI **drafts them in `SCRIPTS.sql` but NEVER executes them** — their promotion to `docs/scripts/` is done by a separate `export-*`, never this loop.
-
 ## Delta 4 — Validation: phase proof + progressive tests
 
 - The phase's **primary proof** is its `Validación de fase`: it demonstrates the **state reached**, not the structure written. Three levels, and the loop never descends one automatically:
@@ -142,14 +138,6 @@ Full gate in [`../CODE-POLICIES.md`](../CODE-POLICIES.md) (§ *Closing review ga
 - **Legacy status line, migrated on write.** A plan carrying the old single-line form (`> Estado: done — YYYY-MM-DD · sesión NNN`) is still **read** as closed; the first time this loop legitimately writes that document, it is rewritten to the two-line form. Compatibility is for reading old plans — every new write uses the normalized contract.
 - **No automatic export**: the artifacts (`SCRIPTS.sql`, `DECISION`, …) stay in the session. Promoting them to `docs/` (scripts, manuals, …) is a separate step via `export-*`.
 
-## Delta 7 — Probe (PoC) tasks
-
-Chassis § *Proof of concept (probe)*, instantiated for execution — for a plan's explicit probe task or a runnable doubt inside a task:
-
-- Seed the question + pass/fail check → run **throwaway code in the session folder** (never the source tree, never committed; DB probe = read-only) → verdict in `CONCLUSIONS`, consequences in `DECISION` (tagged by task) → mark the task with its verdict.
-- A **failed probe does not fail the phase** — it de-risked it: surface it (structured-choice); reshaping the plan goes to `Open questions` + `BACKLOG` (or `/w:plan-refine`).
-- **Promotion**: probe code reaches the sources only as a normal task edit (branch-check + review gate) — never by committing the probe.
-
 ## Sequence
 
 ```
@@ -210,3 +198,8 @@ finalize: CHECKPOINT (+ BACKLOG if something is deferred) + close session + repo
 - A **structural deviation** or a **functional change** exits this loop without converging (§ *Deviation gate*): `CHECKPOINT` + `finalize`, and the work continues in `plan-refine` / `spec-refine`. Same exit when the entry gate finds a structural gap.
 - `Cerrar` (`flow` control, at any time) → `finalize` persists `CHECKPOINT` (and `BACKLOG` only if something remained unexecuted / uncommitted / unapplied), closes the session, reports.
 - Promoting artifacts to `docs/` (via `export-*`) is **always** a later, explicit step outside this loop.
+
+## Conditional modules
+
+- `probe` — probe (PoC) tasks → `../../modules/EXEC-PROBE-TASKS.md`
+- `db` — the DB policy → `../../modules/EXEC-DB-POLICY.md`

package/skills/w/loops/plan-new-loop/LOOP.md

@@ -33,7 +33,7 @@ PLAN
 
 > **slug**: short kebab-case derived from the spec's Requirement — only `[a-z0-9-]`, ≤ ~5 words / ≤ 40 chars. `aw next-number docs/plans` returns JSON (field `next` = `PPP`); the loop builds the full name. To locate plans, glob `docs/plans/PPP-plan-*.md`.
 
-> **Adoption (command mode 4):** an **externally-built plan** (host plan mode, hand-written, another agent) is materialized by the **command** in a single pass — this loop does not run: transcribe + normalize into the Delta 1 schema, `## Origin` = "adopted from <source>" + attribution; then `plan-refine` closes schema gaps (its coherence gate degrades for spec-less plans). See `../../commands/plan-new.md` § *Input resolution*.
+> **Adoption (command mode 4):** an **externally-built plan** (host plan mode, hand-written, another agent) is materialized by the **command** in a single pass — this loop does not run: transcribe + normalize into the Delta 1 schema, `## Origin` = "adopted from <source>" + attribution; then `plan-refine` closes schema gaps (its coherence gate degrades for spec-less plans). See `../../modules/PLAN-INPUT.md` (the command's `input` module).
 
 ## Inherits
 
@@ -124,10 +124,6 @@ A `### Fn` block is a **verifiable state of the system**, never a list of layers
 
 **Granularity is semantic, not mechanical.** A phase earns its place when it leaves a demonstrable state, can be reviewed as a unit, and — when the change carries temporary behavior — moves or retires a simulation. A task is a **coherent unit of purpose** and may touch several files. Naming an edit operation — "create class X", "add method Y", "update the import" — describes a **micro step**: internal to execution, recorded in `CHECKPOINT` when a resume needs it, never a plan entry. `XS–S` stays an orientation of risk and scope; it never mandates splitting a semantic task into mechanical operations.
 
-## Incremental strategy (reference, never a template)
-
-A change spread over consumer, service and data often lands well as: consumer shell → minimal real integration → vertical skeleton → real implementation from the source outwards → hardening → finish. **Reference to adapt, never a mandatory shape.** Backend-only, CLI, batch, library and database-only changes have their own journey, and inserting a layer the project does not have fails the minimality lens. A small change may be **one phase**, when that phase already is a coherent verifiable state.
-
 ## Delta 2 — Gap taxonomy (of "plan")
 
 Replaces the spec gap taxonomy with a planning-oriented one:
@@ -156,33 +152,6 @@ Replaces the spec gap taxonomy with a planning-oriented one:
 
 The chassis' **inline** research specializes: mapping **code/impact** — affected FE/BE/DB components, AS-IS wiring, dependencies. It feeds `Solution` (its AS-IS → TO-BE delta) and `Impacted`. The chassis DB rule applies unchanged (read-only queries into `SCRIPTS.sql`, MCP chosen via a content question when >1 without default).
 
-## Delta 4 — Design SPECs (when the plan includes UI)
-
-The **UI without design SPEC** gap is resolved by **composing** the **`ui-design`** capability (built-in default [`ui-spec`](../../roles/ui-spec/ROLE.md); rebindable via `.workflow/skills.toml`; `off` → degrades to human / `Open questions`):
-
-- It authors **one design SPEC per screen** as a session artifact: `NNN-SPEC-<SLUG>.md` (numbering local to the session — see [`SPEC.md`](../../artifacts/artifacts-design/SPEC.md)).
-- It **derives** from the spec's `## UI spec` section when present (splits it per screen and raises it to executable detail); otherwise it authors from the `Requirement` (design system/theme/ambiguities via *structured-choice*, counts in the batch).
-- The plan's **UI Tasks reference** their SPEC's path — that reference is the **source of truth** — and `plan-exec-loop` reads them as the design reference.
-- It is the chassis' composed-capability resolution mode (next to *research*, *probe* and *human*).
-
-## Delta 5 — Probe (PoC) tasks — de-risk early
-
-Chassis § *Proof of concept (probe)*, instantiated for planning. Two placements:
-
-- **Plan-shaping unknown** (the `Solution` itself depends on the answer) → run the probe **inline now**; the verdict (`CONCLUSIONS`) feeds `Solution` / `Risks / impact`.
-- **Execution-time risk** (a task will build on a risky, runnable assumption) → encode an explicit **probe task**, placed **early** — before the tasks that depend on its verdict; the matching `Risks / impact` entry references it.
-
-## Split gate (multi-plan)
-
-Resolves the **Plan splittable** gap (Delta 2) — the canonical definition for **both** plan loops (`plan-refine-loop` references it, never redefines it). It fires **only on clear signals** (≥2 of: tranches independently executable/deliverable · no shared deps/risk between tranches · different requested moments/priorities · the plan far exceeds S-complexity phases · the user asked for staging); borderline → **one plan, no question**. It can be assessed during decomposition or at the coherence gate, always **before** `Guardar`.
-
-- **The offer** enters the batch as a **content question** (counts in the ≤3): the body shows the proposed cut in the **user's language** — per sibling, a name + slug, a 1-line scope, the phase mapping and the order. Labels: `Dividir en varios planes` (recommended when the signals hold) | `Un solo plan`. Declining marks the gap **exhausted** (no re-offer this run); a free-form answer adjusts the cut. The accepted cut is seeded into `CHECKPOINT` — a resume does **not** re-ask.
-- **Anti-duplicate** (the `create_or_resume` spirit): if sibling plans whose `## Origin` references this same spec/split already exist, the recommended option becomes resuming them (`/w:plan-refine` / `/w:plan-exec` semantics) — never a second set.
-- **On acceptance** — same run, same session (one session per run, one HISTORY row): **all N siblings are elaborated complete** in this run — each gets the full Delta 1 schema and is immediately executable (`plan-exec` runs any plan; a seed without `## Tasks` would break that contract). Context pressure is absorbed by self-regulation (chassis § *Compact / resume*). Numbering: `aw next-number docs/plans` **immediately before each write** — numbers come out consecutive, so every sibling path is known after the first mint.
-- **Sibling contract**: each `## Origin` records the shared source spec + `split (part i/N)` + the **siblings by path** + the order; `## Dependencies` (the existing optional section) carries the inter-plan order — **acyclic and advisory** (`plan-exec` does not enforce it; it only orients what to attack first).
-- **Coherence gate, re-framed**: every spec acceptance criterion traces to **exactly one** sibling — a **complete, disjoint partition**; each sibling's Final behavior block (in `## Solution`) covers its subset; the union covers the spec. Spec-less plans anchor the partition to their own Final behavior block / `Validations`.
-- **Closing action** on the split branch: `Guardar planes` (the single-plan branch keeps `Guardar plan`).
-
 ## Sequence
 
 ```
@@ -229,3 +198,10 @@ finalize: CHECKPOINT persisted (+ BACKLOG only if something is deferred) + close
 - `Cerrar` at any time → `finalize` (persists `CHECKPOINT`; `BACKLOG` only if something is deferred; closes the session, reports).
 
 > **After generating:** the plan can go straight to `plan-exec`, or — if changes arise before executing (new requirements, scope adjustments) — pass through [`plan-refine-loop`](../plan-refine-loop/LOOP.md) (`/w:plan-refine`, auxiliary and **not mandatory**), which refines it in place.
+
+## Conditional modules
+
+- `split` — the canonical multi-plan gate → `../../modules/PLAN-SPLIT-GATE.md`
+- `split` — the incremental journey shape → `../../modules/INCREMENTAL-STRATEGY.md`
+- `ui` — per-screen design SPECs → `../../modules/PLAN-DESIGN-SPECS.md`
+- `probe` — probe (PoC) tasks → `../../modules/PLAN-PROBE-TASKS.md`

package/skills/w/loops/plan-refine-loop/LOOP.md

@@ -91,10 +91,6 @@ Reuses plan-new-loop's gap taxonomy **in full** ([`plan-new-loop`](../plan-new-l
 
 Same as plan-new (maps code/impact: FE/BE/DB components, AS-IS wiring, deps), but **scoped to the delta**: it re-verifies only what the change touches (never re-maps the whole plan). Chassis DB rule unchanged (read-only into `SCRIPTS.sql`, MCP via a question when >1 without default).
 
-## Delta 4 — Design SPECs (when the refine touches UI)
-
-Same mechanism as [`plan-new-loop`](../plan-new-loop/LOOP.md) (§ *Delta 4*: the **`ui-design`** capability → per-screen `NNN-SPEC-<SLUG>.md`, see [`SPEC.md`](../../artifacts/artifacts-design/SPEC.md)), **scoped to the delta**: only the screens **new or changed** by the refine get a design SPEC. The updated SPEC is written in **plan-refine's own session** (each loop manages ITS session's artifacts — it never edits plan-new's) and the plan **re-points** the UI Task reference to the current SPEC. Untouched screens keep their original SPEC.
-
 ## Functional journey map
 
 Before re-shaping phases, the plan must answer what execution would otherwise have to invent. The answers land **inside the existing sections** (`## Solution`, `## Impacted`, the `### Fn` blocks) — this loop adds no plan section beyond its trace:
@@ -106,14 +102,6 @@ Before re-shaping phases, the plan must answer what execution would otherwise ha
 
 > **Bounded research** (Delta 3, sharpened): investigate **only** what is needed to order the phases — journey, boundaries, existing contracts, where the simulation can live, risks that change the plan's design. What `plan-exec` resolves locally is **not** pre-investigated here.
 
-## Simulation lifecycle
-
-**This section applies only when the journey introduces temporary behavior** — a stub, fake, in-memory adapter, controlled fixture or temporary response. None in the change → no `Límite de simulación`, and no artificial phase invented to retire one. When there is, it is planned, never improvised: every simulation declares **purpose · location · the contract it stands for · the phase where it appears · the phase where it moves or disappears · what prevents its accidental selection in a production runtime · the minimum proof needed while it exists**.
-
-- **Explicit over hidden**: `Stub…` / `Fake…`, in-memory adapter, temporary provider or controlled fixture — never a hardcode buried inside production code.
-- **Displacement rule**: each affected phase writes its `Límite de simulación` as normalized prose — `antes <where it is>` → `después <where it lands, or removed>` — never as an implicit assumption.
-- **Removal gate**: the change is not complete while a main-path simulation stays active, a configuration can still select it, the plan does not explain why it remains, or its removal was never validated. Test doubles isolated from the production runtime may stay.
-
 ## Evidence by behavior
 
 Each phase carries proof of behavior, not of structure. Three levels, **chosen** — never descended automatically:
@@ -137,30 +125,6 @@ This loop's instance of the chassis convergence gate — the same one `plan-exec
 - **Evidence** — every phase declares its primary proof, per-layer tests are justified, the same scenario is not duplicated by default, declared risks have evidence or an explicit deferral.
 - **Resumability** — tasks legible enough for a `CHECKPOINT`, intermediate states stable, pending work distinguishable from work already `validada`.
 
-## Replanning executed work
-
-A partially executed plan is replanned **forward**, never rewritten backwards:
-
-- Phases already `validada` **stay as they are**, and their result becomes the initial state of what follows. Completed tasks are not re-written as if they had never happened.
-- Only pending work is re-designed. A `validada` phase that the new shape invalidates gets a **compensating correction** as a new phase — never a silent edit of the closed one.
-- **Return from `plan-exec`** (structural deviation — [`plan-exec-loop`](../plan-exec-loop/LOOP.md) § *Deviation gate*): the deviation enters as a **material gap** of this run. Changed input or output, observable state, public contract, participating components, phase order, simulation boundary or integration strategy — all land in `## Refinement decisions` with what execution already proved.
-- **Legacy plans** (no phase state, micro-task shape) are migrated to the phase contract when this round touches them: completed tasks preserved, micro-tasks grouped by purpose, results and exit conditions derived from the spec and from the plan itself. Functional evidence that neither backs is **not invented** — it becomes an `## Open questions` entry.
-
-## Split gate — refine semantics
-
-The gate itself — signals, offer, anti-duplicate, sibling contract, partition — is defined **once** in [`plan-new-loop`](../plan-new-loop/LOOP.md) § *Split gate (multi-plan)* (this flow's guaranteed load already includes that file); this loop only adds the **in-place semantics** of splitting an existing plan:
-
-- The original plan **keeps its number/path**: it is rewritten **reduced** to its remaining tranche (in place, with confirmation). The extracted tranches become newly minted sibling plans (`aw next-number docs/plans` immediately before each write); their `## Origin` records "split from `docs/plans/PPP-plan-<slug>.md`" + the source spec + the siblings by path.
-- The gate also fires on **partially executed** plans. **Completed tasks (`- [x]`) never move to a sibling** — execution history stays anchored to the original path (plan-exec sessions' `## Origin` keep resolving); only pending work is extracted.
-- The split is recorded in `## Refinement decisions` (what moved where + why); original + siblings together keep the **complete, disjoint partition** of the spec criteria (spec-less: the Delta 2 degradation applies).
-- **Closing action** on the split branch: `Guardar planes` (edit the original reduced + write the extracted siblings); the normal branch keeps `Guardar plan refinado`.
-
-## Compact / resume — PLAN-refine keys
-
-Full mechanism (3 cases, `Compactar`, re-run with `--reopen`) in the chassis (§ *Compact / resume*). PLAN-refine keys: prior-work mark = `## Refinement decisions` **in the plan** (legacy plans may also carry `## Q&A traceability`); re-refine on demand is **first-class** as many times as needed while the flow stays in PLAN.
-
-> **Inter-turn continuity** (chassis, row 2): a flow command opens a "new work line" (new session) — **except re-running the same flow over the same input** (same plan), which does `create_or_resume` (resumes/reopens instead of duplicating).
-
 ## Sequence
 
 ```
@@ -202,3 +166,11 @@ finalize: CHECKPOINT persisted (+ BACKLOG only if something is deferred) + close
 - `Cerrar` at any time → `finalize` (persists `CHECKPOINT`; `BACKLOG` only if something is deferred; closes the session, reports).
 
 > **After refining:** the plan goes to `plan-exec`, which re-checks this same gate on entry — a plan saved here should never bounce back for its structure.
+
+## Conditional modules
+
+- `ui` — design SPECs for changed screens → `../../modules/PLAN-REFINE-DESIGN-SPECS.md`
+- `simulation` — the life of temporary behavior → `../../modules/SIMULATION-LIFECYCLE.md`
+- `replan` — replanning executed work → `../../modules/REPLANNING.md`
+- `split` — splitting a plan in place → `../../modules/PLAN-REFINE-SPLIT.md`
+- `resume` — this loop's compact / resume keys → `../../modules/PLAN-REFINE-KEYS.md`

package/skills/w/loops/quick-loop/LOOP.md

@@ -67,16 +67,6 @@ Read **[`../CHASSIS.md`](../CHASSIS.md)** — the loop's **full engine** — **a
   3. The artifacts (`DECISION`, `SCRIPTS.sql`) **stay in the quick session** as referenceable context for the new session (never migrated).
   4. **SPEC enters live**: after `finalize`, run the *Live transition to SPEC* (draft **only if no spec exists** for this objective; then the loop). **Asymmetry** intact: PLAN can **absorb** the progress (plan-exec picks up the existing working tree); SPEC **restarts** the design cycle and treats the half-done code as context/reference, never as ingested work.
 
-## Continuity across prompts (operating context)
-
-`quick` is where the **continuity rule** ([`../../SKILL.md`](../../SKILL.md) § *Operating context*) shows most clearly. Inside a workspace:
-
-1. `/w:quick "first prompt"` (**command**) → creates session `NNN-<slug>-quick`, starts the loop. Scripts go to **its** `SCRIPTS.sql`.
-2. `"second prompt"` (**no command**, related work) → does **not** create another session: **continues/reopens the most recent one** (from step 1) and appends the new scripts to **that same** `SCRIPTS.sql`.
-3. `/w:quick "third prompt"` (**command** again) → **new** session, new loop.
-
-> The **command** signals "new work line"; a **bare prompt** means "same line" → by default continue/reopen the most recent session (the *last started*). Clearly unrelated → offer choosing (`continuar NNN` | `trabajo nuevo`) or fall to the **no-flow** branch (write into `docs/` by convention + numbering). No workspace → **vanilla** behavior.
-
 ## Sequence
 
 ```
@@ -123,3 +113,7 @@ finalize: CHECKPOINT (AFTER: Pending→Completed) + BACKLOG (only if something i
 - **No export**: nothing goes to `docs/`. Anything worth preserving → promoted separately via `export-*`, or escalated (to SPEC **live** — the line continues in spec-refine already as SPEC flow; to PLAN **deferred**, seed + pointer).
 
 > QUICK's *convergence gate* is **proportional verification-first**: a **short** `Success criteria` seeded at start (not the *absence* of a checklist — its minimal version) — for code, "the change does what the prompt asked + tests/build green"; for analysis/design, a short ratified rubric. Minimal ceremony by design, but **always with the check declared first**.
+
+## Conditional modules
+
+- `resume` — continuity across prompts → `../../modules/PROMPT-CONTINUITY.md`