cairn-ai 0.1.6 → 0.1.8

package/.agents/workflows/cairn-memory.md

@@ -9,7 +9,7 @@ Procedure:
 1. Ensure `MEMORY.md`, `PLAN.md`, `docs/memory`, and `docs/plan`.
 2. Every agent must read the project-root `MEMORY.md` for domain knowledge and repository policy before assigned discovery or update work.
 3. Explore repository facts with focused tools.
-4. Use `explorer` for read-only domain discovery and contradiction checks when available; use `worker` for exact references, command checks, and bounded evidence capture. Tell every delegated agent to read the project-root `MEMORY.md` before work.
+4. Use `explorer` for read-only domain discovery and contradiction checks when available; use `worker` for exact references, command checks, and bounded evidence capture. When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents. Tell every delegated agent and child subagent to read the project-root `MEMORY.md` before work, keep scope, and preserve others' edits.
 5. Write details to `docs/memory/<domain>.md`.
 6. Keep `MEMORY.md` as a short index.
 7. Use the OS locale for user-visible output unless the user asks for another language.

package/.agents/workflows/cairn-plan.md

@@ -12,11 +12,12 @@ Procedure:
 4. Run `node scripts/cairn.mjs toolcheck` when available.
 5. If required LSP/check tools are missing, run `node scripts/cairn.mjs toolcheck --install` or a repository-native install command before accepting a fallback.
 6. Explore before asking the user.
-7. Run complexity triage before applying agent, plugin, or delegated workflow guidance, and record it before mutating files.
-8. Use Light Path by default unless a Heavy Path condition clearly applies.
-9. Use `explorer` for independent read-only discovery, impact analysis, pattern search, and read-only verification when available.
-10. Use `worker` for bounded implementation or verification tasks with clear file ownership. Tell every delegated agent to read the project-root `MEMORY.md` before work. Keep urgent blocking work local.
-11. Write `docs/plan/<topic>.md`.
-12. Add the topic link and status to `PLAN.md`.
-13. Put the selected Light/Heavy Path, checked Heavy Path signals, tool readiness, module acceptance verification, and surface integration verification on every module slice.
-14. Use the OS locale for user-visible output unless the user asks for another language.
+7. Understand the whole work and affected surfaces first, then classify it into small executable tasks. If a task is still too broad to verify cleanly, split it into sub-tasks.
+8. Run complexity triage before applying agent, plugin, or delegated workflow guidance, and record it before mutating files.
+9. Use Light Path by default unless a Heavy Path condition clearly applies.
+10. Use `explorer` for independent read-only discovery, impact analysis, pattern search, and read-only verification when available.
+11. Use `worker` for bounded implementation or verification tasks with clear file ownership. When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents. Tell every delegated agent and child subagent to read the project-root `MEMORY.md` before work, keep scope, and preserve others' edits. Keep urgent blocking work local.
+12. Write `docs/plan/<topic>.md`.
+13. Add the topic link and status to `PLAN.md`.
+14. Put the selected Light/Heavy Path, checked Heavy Path signals, tool readiness, module acceptance verification, and surface integration verification on every module task.
+15. Use the OS locale for user-visible output unless the user asks for another language.

package/.agents/workflows/cairn-review.md

@@ -6,9 +6,9 @@ Goal: review completed work against plan, memory, policy, and evidence.
 
 Procedure:
 
-1. Every agent must read the project-root `MEMORY.md` for domain knowledge and repository policy before review work, then read the plan slice and memory inputs.
+1. Every agent must read the project-root `MEMORY.md` for domain knowledge and repository policy before review work, then read the plan task and memory inputs.
 2. Check changed files, tool readiness evidence, and verification evidence.
-3. Delegate independent read-only review to `explorer` when it materially improves speed or quality; use `worker` only for scoped verification reruns or QA artifacts. Tell every delegated agent to read the project-root `MEMORY.md` before work.
+3. Delegate independent read-only review to `explorer` when it materially improves speed or quality; use `worker` only for scoped verification reruns or QA artifacts. When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents. Tell every delegated agent and child subagent to read the project-root `MEMORY.md` before work, keep scope, and preserve others' edits.
 4. Report findings first, ordered by severity, with file and line references. If no issues are found, say so clearly and mention remaining test gaps or residual risk.
 5. Record review evidence in `docs/plan/<topic>.md`.
 6. Use the OS locale for user-visible output unless the user asks for another language.

package/.agents/workflows/cairn-work.md

@@ -2,20 +2,21 @@
 
 Use the `cairn-work` skill.
 
-Goal: execute the next incomplete module slice in `PLAN.md` without bypassing the recorded complexity triage.
+Goal: execute the next incomplete module task in `PLAN.md` without bypassing the recorded complexity triage.
 
 Procedure:
 
-1. Every agent must read the project-root `MEMORY.md` for domain knowledge and repository policy before selecting or executing a slice.
+1. Every agent must read the project-root `MEMORY.md` for domain knowledge and repository policy before selecting or executing a task.
 2. Read `PLAN.md`, the detailed plan, and relevant memory notes.
 3. Read the Codex or Claude model guidance recorded in the plan.
-4. Select one small module slice.
+4. Select one small module task.
 5. Confirm required tool readiness. If a required LSP/check tool is missing, run `node scripts/cairn.mjs toolcheck --install` or a repository-native install command before accepting a fallback.
 6. Confirm the complexity triage and selected Light/Heavy Path recorded in the plan. If missing, update the plan before mutating files.
 7. For Light Path, implement directly or use one bounded `worker`, then run focused verification.
 8. For Heavy Path, follow the plan's staged implementation and review gates without role-mapping shortcuts.
-9. When delegating implementation to `worker`, state the file ownership clearly, tell every delegated agent to read the project-root `MEMORY.md` before work, and tell the worker not to revert others' edits.
+9. When delegating implementation to `worker`, state the file ownership clearly. When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents. Tell every delegated agent and child subagent to read the project-root `MEMORY.md` before work, keep scope, and preserve others' edits.
 10. Re-run module acceptance verification.
 11. Re-run surface integration verification.
 12. Record tool readiness and verification evidence in `docs/plan/<topic>.md` and update `PLAN.md`.
-13. Use the OS locale for user-visible output unless the user asks for another language.
+13. If the user asks a side question, status question, or narrow clarification while this task is still active, answer it briefly and then resume the previous active work unless the user explicitly asks to pause, stop, or switch tasks.
+14. Use the OS locale for user-visible output unless the user asks for another language.

package/.claude/commands/cairn-review.md

@@ -1,5 +1,5 @@
 # cairn-review
 
-Review completed slices against memory, plan, policy, tool readiness, and evidence.
+Review completed tasks against memory, plan, policy, tool readiness, and evidence.
 
 Follow plugin command `commands/cairn-review.md`.

package/.claude/commands/cairn-work.md

@@ -1,5 +1,7 @@
 # cairn-work
 
-Start by reading the project-root `MEMORY.md`, then execute the next incomplete module slice only after the plan records Light/Heavy Path complexity triage, tool readiness, module acceptance verification, and surface integration verification.
+Start by reading the project-root `MEMORY.md`, then execute the next incomplete module task only after the plan records Light/Heavy Path complexity triage, tool readiness, module acceptance verification, and surface integration verification.
+
+If the user asks a side question, status question, or narrow clarification while a task is still active, answer it briefly and then resume the previous active work unless the user explicitly asks to pause, stop, or switch tasks.
 
 Follow plugin command `commands/cairn-work.md`.

package/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "cairn",
-  "version": "0.1.6",
+  "version": "0.1.8",
   "description": "Token-efficient agent harness with tool readiness for Codex, Claude Code, and Antigravity.",
   "author": {
     "name": "WkNam"
@@ -9,7 +9,7 @@
   "interface": {
     "displayName": "Cairn",
     "shortDescription": "Memory, planning, tool readiness, and two-gate verification harness for agent teams.",
-    "longDescription": "Cairn brings a LazyCodex-inspired flow to Codex, Claude Code, and Antigravity while reducing token use through repository-local MEMORY.md, PLAN.md, domain notes, tool readiness checks, module-slice execution, and two-gate verification.",
+    "longDescription": "Cairn brings a LazyCodex-inspired flow to Codex, Claude Code, and Antigravity while reducing token use through repository-local MEMORY.md, PLAN.md, domain notes, tool readiness checks, module-task execution, and two-gate verification.",
     "developerName": "WkNam",
     "category": "Productivity",
     "capabilities": [
@@ -25,8 +25,11 @@
       "At the start of each task, every agent must read the project-root MEMORY.md for domain knowledge and repository policy before planning, implementation, review, or delegation.",
       "Run cairn-toolcheck before planning implementation work.",
       "Before non-trivial implementation, run cairn-plan for this task without asking me for discoverable facts.",
+      "When planning, first understand the whole work, then classify it into small executable tasks; if a task needs further breakdown, split it into sub-tasks.",
+      "When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents; every child subagent must read MEMORY.md, keep scope, and preserve others' edits.",
       "Record Light Path or Heavy Path complexity triage in docs/plan/<topic>.md before mutating files.",
-      "Run cairn-work from the current PLAN.md and pass dry-run/check, module acceptance, and surface integration gates."
+      "Run cairn-work from the current PLAN.md and pass dry-run/check, module acceptance, and surface integration gates.",
+      "If I ask a side question, status question, or narrow clarification while a plan or task is still active, answer it briefly and then resume the previous active work unless I explicitly ask you to pause, stop, or switch tasks."
     ]
   }
 }

package/README.de.md

@@ -4,12 +4,12 @@ Cairn ist ein token-effizientes Multi-Agent-Harness-Plugin für Codex, Claude Co
 
 [English](README.md)
 
-Die Grundidee ist, nützliches agent-harness-Verhalten beizubehalten: hooks, persistent state, explicit planning, focused delegation und stop-time guards. Cairn macht wiederholte TDD verification loops nicht zum Standard. Stattdessen teilt es Arbeit in kleine module slices und belegt jeden slice mit zwei verification gates.
+Die Grundidee ist, nützliches agent-harness-Verhalten beizubehalten: hooks, persistent state, explicit planning, focused delegation und stop-time guards. Cairn macht wiederholte TDD verification loops nicht zum Standard. Stattdessen teilt es Arbeit in kleine module tasks und belegt jeden task mit zwei verification gates.
 
 1. Module acceptance verification: belegt den geänderten module contract.
 2. Surface integration verification: belegt Verhalten über die echte surface, etwa CLI, HTTP, browser oder file artifacts.
 
-Bevor ein slice externen Zustand verändert, zeichnet Cairn den nächstliegenden dry-run oder check mode auf und führt ihn aus. Verification ist begrenzt: Jeder slice erhält standardmäßig zwei verification passes; danach zeichnet der agent einen blocker auf oder teilt den slice, statt eine offene Schleife fortzusetzen.
+Bevor ein task externen Zustand verändert, zeichnet Cairn den nächstliegenden dry-run oder check mode auf und führt ihn aus. Verification ist begrenzt: Jeder task erhält standardmäßig zwei verification passes; danach zeichnet der agent einen blocker auf oder teilt den task, statt eine offene Schleife fortzusetzen.
 
 Cairn behandelt tool readiness ebenfalls als Teil der Arbeit. LSP, typecheck, lint, dry-run und verification tools werden gegen den repository stack geprüft. Fehlt ein erforderliches tool, versucht Cairn eine project-local oder repository-native installation, bevor ein fallback akzeptiert wird.
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - Migrations und database changes verwenden vor write/apply commands `--pretend`, dry-run, schema diff, rollback feasibility checks oder das nächstliegende repository-native equivalent.
 - Package, release, infrastructure, deployment, code generation und formatting work verwenden, wenn verfügbar, check, plan, diff, validate oder dry-run modes, bevor Zustand verändert wird.
 - Wenn kein dry-run existiert, hält der plan dies fest und wählt den kleinsten reversiblen command oder test artifact.
-- Wenn ein verification gate fehlschlägt, diagnostiziert Cairn einmal, verkleinert oder teilt den slice und führt beide gates erneut aus.
-- Nach zwei fehlgeschlagenen verification passes für denselben slice zeichnet Cairn den blocker in `docs/plan/<topic>.md` auf, statt eine wiederholte Schleife fortzusetzen.
+- Wenn ein verification gate fehlschlägt, diagnostiziert Cairn einmal, verkleinert oder teilt den task und führt beide gates erneut aus.
+- Nach zwei fehlgeschlagenen verification passes für denselben task zeichnet Cairn den blocker in `docs/plan/<topic>.md` auf, statt eine wiederholte Schleife fortzusetzen.
 
 ## Model Guidance
 
 Cairn wendet model-specific adjustment nur auf Claude-family und Codex-family models an.
 
 - Claude-family: geeignet für long context, policy interpretation und plan/evidence review.
-- Codex-family: geeignet für small implementation slices, explicit file edits, command-based verification und bounded `worker` tasks.
+- Codex-family: geeignet für small implementation tasks, explicit file edits, command-based verification und bounded `worker` tasks.
 
 Details stehen in `docs/model-guidance/README.md`, `docs/model-guidance/claude.md` und `docs/model-guidance/codex.md`.
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck`: erkennt repository stacks und prüft oder installiert required LSP und verification tools.
 - `cairn-memory`: erkundet domain knowledge und aktualisiert `MEMORY.md`.
 - `cairn-plan`: erstellt einen decision-complete plan unter `docs/plan/`.
-- `cairn-work`: führt den nächsten module slice im aktuellen `PLAN.md` mit zwei verification gates aus.
-- `cairn-review`: prüft completed slices gegen plan, memory und evidence.
+- `cairn-work`: führt den nächsten module task im aktuellen `PLAN.md` mit zwei verification gates aus.
+- `cairn-review`: prüft completed tasks gegen plan, memory und evidence.
 
 `install` und `upgrade` erstellen `*.cairn-backup-*` backups, bevor `~/.codex/config.toml` verändert wird. Das source plugin manifest bleibt validator-friendly; nur die installed cache copy erhält ein `hooks` field zur Aktivierung von Codex hooks.
 

package/README.es.md

@@ -4,12 +4,12 @@ Cairn es un plugin de harness multi-agente eficiente en tokens para Codex, Claud
 
 [English](README.md)
 
-La idea central es conservar comportamientos útiles de agent harness: hooks, estado persistente, planificación explícita, delegación enfocada y guardas al detenerse. Cairn no hace que los bucles repetidos de verificación TDD sean el comportamiento predeterminado. En su lugar, divide el trabajo en pequeños module slices y prueba cada slice con dos verification gates.
+La idea central es conservar comportamientos útiles de agent harness: hooks, estado persistente, planificación explícita, delegación enfocada y guardas al detenerse. Cairn no hace que los bucles repetidos de verificación TDD sean el comportamiento predeterminado. En su lugar, divide el trabajo en pequeños module tasks y prueba cada task con dos verification gates.
 
 1. Module acceptance verification: prueba el contrato del módulo modificado.
 2. Surface integration verification: prueba el comportamiento a través de una superficie real, como CLI, HTTP, navegador o artefactos de archivo.
 
-Antes de que un slice modifique estado externo, Cairn registra y ejecuta el dry-run o check mode más cercano disponible. La verificación está acotada: cada slice tiene dos verification passes por defecto; después, el agente registra un blocker o divide el slice en lugar de continuar un bucle abierto.
+Antes de que un task modifique estado externo, Cairn registra y ejecuta el dry-run o check mode más cercano disponible. La verificación está acotada: cada task tiene dos verification passes por defecto; después, el agente registra un blocker o divide el task en lugar de continuar un bucle abierto.
 
 Cairn también trata la preparación de herramientas como parte del trabajo. LSP, typecheck, lint, dry-run y verification tools se revisan contra el stack del repositorio. Si falta una herramienta requerida, Cairn intenta una instalación project-local o repository-native antes de aceptar un fallback.
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - Migrations y database changes usan `--pretend`, dry-run, schema diff, rollback feasibility checks o el equivalente repository-native más cercano antes de comandos write/apply.
 - Package, release, infrastructure, deployment, code generation y formatting work usan check, plan, diff, validate o dry-run modes antes de mutar estado cuando están disponibles.
 - Si no existe dry-run, el plan registra ese hecho y elige el comando reversible más pequeño o el artefacto de prueba disponible.
-- Si falla un verification gate, Cairn diagnostica una vez, reduce o divide el slice y vuelve a ejecutar ambos gates.
-- Después de dos verification passes fallidos para el mismo slice, Cairn registra el blocker en `docs/plan/<topic>.md` en lugar de continuar un bucle repetido.
+- Si falla un verification gate, Cairn diagnostica una vez, reduce o divide el task y vuelve a ejecutar ambos gates.
+- Después de dos verification passes fallidos para el mismo task, Cairn registra el blocker en `docs/plan/<topic>.md` en lugar de continuar un bucle repetido.
 
 ## Model Guidance
 
 Cairn solo aplica ajustes específicos de modelo a modelos Claude-family y Codex-family.
 
 - Claude-family: útil para long context, policy interpretation y plan/evidence review.
-- Codex-family: útil para small implementation slices, explicit file edits, command-based verification y bounded `worker` tasks.
+- Codex-family: útil para small implementation tasks, explicit file edits, command-based verification y bounded `worker` tasks.
 
 La guía detallada vive en `docs/model-guidance/README.md`, `docs/model-guidance/claude.md` y `docs/model-guidance/codex.md`.
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck`: detecta repository stacks y revisa o instala required LSP y verification tools.
 - `cairn-memory`: explora domain knowledge y actualiza `MEMORY.md`.
 - `cairn-plan`: crea un decision-complete plan bajo `docs/plan/`.
-- `cairn-work`: ejecuta el siguiente module slice del `PLAN.md` actual con dos verification gates.
-- `cairn-review`: revisa completed slices contra plan, memory y evidence.
+- `cairn-work`: ejecuta el siguiente module task del `PLAN.md` actual con dos verification gates.
+- `cairn-review`: revisa completed tasks contra plan, memory y evidence.
 
 `install` y `upgrade` crean backups `*.cairn-backup-*` antes de modificar `~/.codex/config.toml`. El source plugin manifest permanece validator-friendly; solo la copia instalada en cache recibe un campo `hooks` para activar Codex hooks.
 

package/README.fr.md

@@ -4,12 +4,12 @@ Cairn est un plugin de harness multi-agent économe en tokens pour Codex, Claude
 
 [English](README.md)
 
-L'idée centrale est de conserver des comportements utiles d'agent harness : hooks, état persistant, planification explicite, délégation ciblée et gardes au moment de l'arrêt. Cairn ne fait pas des boucles répétées de vérification TDD le comportement par défaut. Il découpe plutôt le travail en petits module slices et prouve chaque slice avec deux verification gates.
+L'idée centrale est de conserver des comportements utiles d'agent harness : hooks, état persistant, planification explicite, délégation ciblée et gardes au moment de l'arrêt. Cairn ne fait pas des boucles répétées de vérification TDD le comportement par défaut. Il découpe plutôt le travail en petits module tasks et prouve chaque task avec deux verification gates.
 
 1. Module acceptance verification : prouve le contrat du module modifié.
 2. Surface integration verification : prouve le comportement via une surface réelle, comme CLI, HTTP, navigateur ou artefacts de fichiers.
 
-Avant qu'un slice ne modifie un état externe, Cairn enregistre et exécute le dry-run ou check mode disponible le plus proche. La vérification est bornée : chaque slice reçoit deux verification passes par défaut, puis l'agent enregistre un blocker ou divise le slice au lieu de continuer une boucle ouverte.
+Avant qu'un task ne modifie un état externe, Cairn enregistre et exécute le dry-run ou check mode disponible le plus proche. La vérification est bornée : chaque task reçoit deux verification passes par défaut, puis l'agent enregistre un blocker ou divise le task au lieu de continuer une boucle ouverte.
 
 Cairn traite aussi la préparation des outils comme une partie du travail. LSP, typecheck, lint, dry-run et verification tools sont vérifiés par rapport au stack du repository. Si un outil requis manque, Cairn tente une installation project-local ou repository-native avant d'accepter un fallback.
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - Migrations et database changes utilisent `--pretend`, dry-run, schema diff, rollback feasibility checks ou l'équivalent repository-native le plus proche avant les commandes write/apply.
 - Package, release, infrastructure, deployment, code generation et formatting work utilisent check, plan, diff, validate ou dry-run modes avant de modifier l'état quand c'est disponible.
 - Si aucun dry-run n'existe, le plan l'enregistre et choisit la plus petite commande réversible ou le test artifact disponible.
-- Si une verification gate échoue, Cairn diagnostique une fois, réduit ou divise le slice, puis relance les deux gates.
-- Après deux verification passes échouées pour le même slice, Cairn enregistre le blocker dans `docs/plan/<topic>.md` au lieu de continuer une boucle répétée.
+- Si une verification gate échoue, Cairn diagnostique une fois, réduit ou divise le task, puis relance les deux gates.
+- Après deux verification passes échouées pour le même task, Cairn enregistre le blocker dans `docs/plan/<topic>.md` au lieu de continuer une boucle répétée.
 
 ## Model Guidance
 
 Cairn applique des ajustements spécifiques uniquement aux modèles Claude-family et Codex-family.
 
 - Claude-family : utile pour long context, policy interpretation et plan/evidence review.
-- Codex-family : utile pour small implementation slices, explicit file edits, command-based verification et bounded `worker` tasks.
+- Codex-family : utile pour small implementation tasks, explicit file edits, command-based verification et bounded `worker` tasks.
 
 Les détails se trouvent dans `docs/model-guidance/README.md`, `docs/model-guidance/claude.md` et `docs/model-guidance/codex.md`.
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck` : détecte repository stacks et vérifie ou installe required LSP et verification tools.
 - `cairn-memory` : explore domain knowledge et met à jour `MEMORY.md`.
 - `cairn-plan` : crée un decision-complete plan sous `docs/plan/`.
-- `cairn-work` : exécute le prochain module slice du `PLAN.md` courant avec deux verification gates.
-- `cairn-review` : relit completed slices par rapport au plan, memory et evidence.
+- `cairn-work` : exécute le prochain module task du `PLAN.md` courant avec deux verification gates.
+- `cairn-review` : relit completed tasks par rapport au plan, memory et evidence.
 
 `install` et `upgrade` créent des backups `*.cairn-backup-*` avant de modifier `~/.codex/config.toml`. Le source plugin manifest reste validator-friendly ; seule la copie installée en cache reçoit un champ `hooks` pour activer les Codex hooks.
 

package/README.ja.md

@@ -4,12 +4,12 @@ Cairn は Codex、Claude Code、Antigravity のためのトークン効率の高
 
 [English](README.md)
 
-Cairn の基本方針は、hooks、永続状態、明示的な planning、focused delegation、stop-time guards といった有用な agent harness の動作を残すことです。Cairn は反復的な TDD verification loop を既定にしません。代わりに作業を小さな module slice に分け、2 つの verification gate で証明します。
+Cairn の基本方針は、hooks、永続状態、明示的な planning、focused delegation、stop-time guards といった有用な agent harness の動作を残すことです。Cairn は反復的な TDD verification loop を既定にしません。代わりに作業を小さな module task に分け、2 つの verification gate で証明します。
 
 1. Module acceptance verification: 変更された module contract を証明します。
 2. Surface integration verification: CLI、HTTP、browser、file artifacts など実際の surface を通して動作を証明します。
 
-slice が外部状態を変更する前に、Cairn は最も近い dry-run または check mode を記録して実行します。verification は制限されます。既定では各 slice は 2 回の verification pass を持ち、その後は反復 loop を続けず blocker を記録するか slice を分割します。
+task が外部状態を変更する前に、Cairn は最も近い dry-run または check mode を記録して実行します。verification は制限されます。既定では各 task は 2 回の verification pass を持ち、その後は反復 loop を続けず blocker を記録するか task を分割します。
 
 Cairn は tool readiness も作業の一部として扱います。LSP、typecheck、lint、dry-run、verification tools を repository stack に照らして確認します。必要な tool がない場合、fallback を受け入れる前に project-local または repository-native installation を試みます。
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - Migrations と database changes は write/apply commands の前に `--pretend`、dry-run、schema diff、rollback feasibility checks、または最も近い repository-native equivalent を使います。
 - Package、release、infrastructure、deployment、code generation、formatting work は可能な場合、状態変更の前に check、plan、diff、validate、dry-run modes を使います。
 - dry-run がない場合、plan はその事実を記録し、最小で reversible な command または test artifact を選びます。
-- verification gate が失敗した場合、Cairn は一度診断し、slice を縮小または分割して両方の gate を再実行します。
-- 同じ slice で 2 回の verification pass が失敗した場合、反復 loop を続けず `docs/plan/<topic>.md` に blocker を記録します。
+- verification gate が失敗した場合、Cairn は一度診断し、task を縮小または分割して両方の gate を再実行します。
+- 同じ task で 2 回の verification pass が失敗した場合、反復 loop を続けず `docs/plan/<topic>.md` に blocker を記録します。
 
 ## Model Guidance
 
 Cairn は Claude-family と Codex-family models にのみ model-specific adjustment を適用します。
 
 - Claude-family: long context、policy interpretation、plan/evidence review に有用です。
-- Codex-family: small implementation slices、explicit file edits、command-based verification、bounded `worker` tasks に有用です。
+- Codex-family: small implementation tasks、explicit file edits、command-based verification、bounded `worker` tasks に有用です。
 
 詳細は `docs/model-guidance/README.md`、`docs/model-guidance/claude.md`、`docs/model-guidance/codex.md` にあります。
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck`: repository stacks を検出し、required LSP と verification tools を確認または install します。
 - `cairn-memory`: domain knowledge を探索し `MEMORY.md` を更新します。
 - `cairn-plan`: `docs/plan/` に decision-complete plan を作成します。
-- `cairn-work`: 現在の `PLAN.md` の次の module slice を実行し、2 つの verification gates を取得します。
-- `cairn-review`: plan、memory、evidence に照らして completed slices を review します。
+- `cairn-work`: 現在の `PLAN.md` の次の module task を実行し、2 つの verification gates を取得します。
+- `cairn-review`: plan、memory、evidence に照らして completed tasks を review します。
 
 `install` と `upgrade` は `~/.codex/config.toml` を変更する前に `*.cairn-backup-*` backups を作ります。source plugin manifest は validator-friendly のままにし、Codex hooks を有効化する `hooks` field は installed cache copy にだけ追加されます。
 

package/README.ko.md

@@ -4,12 +4,12 @@ Cairn은 Codex, Claude Code, Antigravity를 위한 토큰 효율적인 멀티에
 
 [English](README.md)
 
-핵심 아이디어는 hooks, 지속 상태, 명시적 계획, 집중 위임, 종료 시점 가드 같은 유용한 에이전트 하네스 동작을 유지하는 것입니다. Cairn은 반복적인 TDD 검증 루프를 기본값으로 삼지 않습니다. 대신 작업을 작은 모듈 조각으로 나누고 두 가지 검증 게이트로 증명합니다.
+핵심 아이디어는 hooks, 지속 상태, 명시적 계획, 집중 위임, 종료 시점 가드 같은 유용한 에이전트 하네스 동작을 유지하는 것입니다. Cairn은 반복적인 TDD 검증 루프를 기본값으로 삼지 않습니다. 대신 작업을 작은 모듈 작업으로 나누고 두 가지 검증 게이트로 증명합니다.
 
 1. 모듈 수용 검증: 변경된 모듈 계약을 증명합니다.
 2. 표면 통합 검증: CLI, HTTP, 브라우저, 파일 산출물 같은 실제 표면을 통해 동작을 증명합니다.
 
-외부 상태를 변경할 수 있는 조각은 실행 전에 가장 가까운 dry-run 또는 check 모드를 기록하고 실행합니다. 검증은 제한됩니다. 기본적으로 각 조각은 두 번의 검증 패스를 가지며, 이후에는 반복 루프를 계속하지 않고 blocker를 기록하거나 조각을 더 작게 나눕니다.
+외부 상태를 변경할 수 있는 작업은 실행 전에 가장 가까운 dry-run 또는 check 모드를 기록하고 실행합니다. 검증은 제한됩니다. 기본적으로 각 작업은 두 번의 검증 패스를 가지며, 이후에는 반복 루프를 계속하지 않고 blocker를 기록하거나 작업을 sub-task로 나눕니다.
 
 Cairn은 도구 준비 상태도 작업의 일부로 다룹니다. LSP, typecheck, lint, dry-run, 검증 도구를 저장소 스택 기준으로 확인하고, 필수 도구가 없으면 fallback을 받아들이기 전에 프로젝트 로컬 또는 저장소 네이티브 설치를 시도합니다.
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - 마이그레이션과 데이터베이스 변경은 write/apply 명령 전에 `--pretend`, dry-run, schema diff, rollback 가능성 검사, 또는 가장 가까운 저장소 네이티브 방식을 사용합니다.
 - 패키지, 릴리스, 인프라, 배포, 코드 생성, 포매팅 작업은 가능한 경우 상태 변경 전에 check, plan, diff, validate, dry-run 모드를 사용합니다.
 - dry-run이 없으면 계획에 그 사실을 기록하고 가장 작고 되돌릴 수 있는 명령 또는 테스트 산출물을 선택합니다.
-- 검증 게이트가 실패하면 Cairn은 한 번 진단하고 조각을 줄이거나 나눈 뒤 두 게이트를 다시 실행합니다.
-- 같은 조각에서 두 번의 검증 패스가 실패하면 반복 루프를 계속하지 않고 `docs/plan/<topic>.md`에 blocker를 기록합니다.
+- 검증 게이트가 실패하면 Cairn은 한 번 진단하고 작업을 줄이거나 sub-task로 나눈 뒤 두 게이트를 다시 실행합니다.
+- 같은 작업에서 두 번의 검증 패스가 실패하면 반복 루프를 계속하지 않고 `docs/plan/<topic>.md`에 blocker를 기록합니다.
 
 ## Model Guidance
 
 Cairn은 Claude-family와 Codex-family 모델에만 모델별 조정을 적용합니다.
 
 - Claude-family: 긴 컨텍스트, 정책 해석, 계획/증거 검토에 유용합니다.
-- Codex-family: 작은 구현 조각, 명시적 파일 편집, command 기반 검증, 제한된 `worker` 작업에 유용합니다.
+- Codex-family: 작은 구현 작업, 명시적 파일 편집, command 기반 검증, 제한된 `worker` 작업에 유용합니다.
 
 자세한 가이드는 `docs/model-guidance/README.md`, `docs/model-guidance/claude.md`, `docs/model-guidance/codex.md`에 있습니다.
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck`: 저장소 스택을 감지하고 필요한 LSP와 검증 도구를 확인하거나 설치합니다.
 - `cairn-memory`: 도메인 지식을 탐색하고 `MEMORY.md`를 갱신합니다.
 - `cairn-plan`: `docs/plan/` 아래에 decision-complete plan을 만듭니다.
-- `cairn-work`: 현재 `PLAN.md`의 다음 모듈 조각을 실행하고 두 검증 게이트를 확보합니다.
-- `cairn-review`: 계획, memory, 증거를 기준으로 완료된 조각을 검토합니다.
+- `cairn-work`: 현재 `PLAN.md`의 다음 모듈 작업을 실행하고 두 검증 게이트를 확보합니다.
+- `cairn-review`: 계획, memory, 증거를 기준으로 완료된 작업을 검토합니다.
 
 `install`과 `upgrade`는 `~/.codex/config.toml`을 수정하기 전에 `*.cairn-backup-*` 백업을 만듭니다. 소스 플러그인 manifest는 validator-friendly 상태를 유지하고, 설치된 cache copy에만 Codex hook 활성화를 위한 `hooks` field가 추가됩니다.
 

package/README.md

@@ -12,12 +12,12 @@ Cairn is a token-efficient multi-agent harness plugin for Codex, Claude Code, an
 - [Deutsch](README.de.md)
 - [Português](README.pt.md)
 
-The core idea is to keep useful agent-harness behavior: hooks, persistent state, explicit planning, focused delegation, and stop-time guards. Cairn does not make repeated TDD verification loops the default. Instead, it splits work into small module slices and proves each slice with two verification gates.
+The core idea is to keep useful agent-harness behavior: hooks, persistent state, explicit planning, focused delegation, and stop-time guards. Cairn does not make repeated TDD verification loops the default. Instead, it splits work into small module tasks and proves each task with two verification gates.
 
 1. Module acceptance verification: proves the changed module contract.
 2. Surface integration verification: proves behavior through the real surface, such as CLI, HTTP, browser, or file artifacts.
 
-Before a slice mutates external state, Cairn records and runs the closest available dry-run or check mode. Verification is bounded: each slice gets two verification passes by default, then the agent records a blocker or splits the slice instead of continuing an open-ended loop.
+Before a task mutates external state, Cairn records and runs the closest available dry-run or check mode. Verification is bounded: each task gets two verification passes by default, then the agent records a blocker or splits it into sub-tasks instead of continuing an open-ended loop.
 
 Cairn also treats tool readiness as part of the work. LSP, typecheck, lint, dry-run, and verification tools are checked against the repository stack. If a required tool is missing, Cairn attempts project-local or repository-native installation before accepting a fallback.
 
@@ -55,15 +55,15 @@ cairn toolcheck --install
 - Migrations and database changes use `--pretend`, dry-run, schema diff, rollback feasibility checks, or the closest repository-native equivalent before write/apply commands.
 - Package, release, infrastructure, deployment, code generation, and formatting work use check, plan, diff, validate, or dry-run modes before mutating state when available.
 - If no dry-run exists, the plan records that fact and selects the smallest reversible command or test artifact available.
-- If a verification gate fails, Cairn diagnoses once, shrinks or splits the slice, and reruns both gates.
-- After two failed verification passes for the same slice, Cairn records the blocker in `docs/plan/<topic>.md` instead of continuing a repeated loop.
+- If a verification gate fails, Cairn diagnoses once, shrinks the task or splits it into sub-tasks, and reruns both gates.
+- After two failed verification passes for the same task, Cairn records the blocker in `docs/plan/<topic>.md` instead of continuing a repeated loop.
 
 ## Model Guidance
 
 Cairn only applies model-specific adjustment to Claude-family and Codex-family models.
 
 - Claude-family: useful for long context, policy interpretation, and plan/evidence review.
-- Codex-family: useful for small implementation slices, explicit file edits, command-based verification, and bounded `worker` tasks.
+- Codex-family: useful for small implementation tasks, explicit file edits, command-based verification, and bounded `worker` tasks.
 
 Detailed guidance lives in `docs/model-guidance/README.md`, `docs/model-guidance/claude.md`, and `docs/model-guidance/codex.md`.
 
@@ -109,8 +109,8 @@ cairn toolcheck
 - `cairn toolcheck`: detects repository stacks and checks or installs required LSP and verification tools.
 - `cairn-memory`: explores domain knowledge and updates `MEMORY.md`.
 - `cairn-plan`: creates a decision-complete plan under `docs/plan/`.
-- `cairn-work`: executes the next module slice in the current `PLAN.md` with two verification gates.
-- `cairn-review`: reviews completed slices against plan, memory, and evidence.
+- `cairn-work`: executes the next module task in the current `PLAN.md` with two verification gates.
+- `cairn-review`: reviews completed tasks against plan, memory, and evidence.
 
 Install and upgrade create `*.cairn-backup-*` backups before modifying `~/.codex/config.toml`. The source plugin manifest stays validator-friendly; only the installed cache copy gets a `hooks` field to activate Codex hooks.
 

package/README.pt.md

@@ -4,12 +4,12 @@ Cairn e um plugin de harness multiagente eficiente em tokens para Codex, Claude
 
 [English](README.md)
 
-A ideia central e manter comportamentos uteis de agent harness: hooks, estado persistente, planejamento explicito, delegacao focada e guardas no momento de parada. Cairn nao torna loops repetidos de verificacao TDD o padrao. Em vez disso, divide o trabalho em pequenos module slices e prova cada slice com dois verification gates.
+A ideia central e manter comportamentos uteis de agent harness: hooks, estado persistente, planejamento explicito, delegacao focada e guardas no momento de parada. Cairn nao torna loops repetidos de verificacao TDD o padrao. Em vez disso, divide o trabalho em pequenos module tasks e prova cada task com dois verification gates.
 
 1. Module acceptance verification: prova o contrato do modulo alterado.
 2. Surface integration verification: prova o comportamento por uma superficie real, como CLI, HTTP, navegador ou artefatos de arquivo.
 
-Antes que um slice altere estado externo, Cairn registra e executa o dry-run ou check mode mais proximo disponivel. A verificacao e limitada: cada slice recebe duas verification passes por padrao; depois, o agente registra um blocker ou divide o slice em vez de continuar um loop aberto.
+Antes que um task altere estado externo, Cairn registra e executa o dry-run ou check mode mais proximo disponivel. A verificacao e limitada: cada task recebe duas verification passes por padrao; depois, o agente registra um blocker ou divide o task em vez de continuar um loop aberto.
 
 Cairn tambem trata tool readiness como parte do trabalho. LSP, typecheck, lint, dry-run e verification tools sao verificados contra o stack do repositorio. Se uma ferramenta obrigatoria estiver ausente, Cairn tenta uma instalacao project-local ou repository-native antes de aceitar um fallback.
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - Migrations e database changes usam `--pretend`, dry-run, schema diff, rollback feasibility checks ou o equivalente repository-native mais proximo antes de write/apply commands.
 - Package, release, infrastructure, deployment, code generation e formatting work usam check, plan, diff, validate ou dry-run modes antes de alterar estado quando disponiveis.
 - Se nao houver dry-run, o plan registra esse fato e escolhe o menor command reversivel ou test artifact disponivel.
-- Se um verification gate falhar, Cairn diagnostica uma vez, reduz ou divide o slice e executa novamente os dois gates.
-- Depois de duas verification passes falhas para o mesmo slice, Cairn registra o blocker em `docs/plan/<topic>.md` em vez de continuar um loop repetido.
+- Se um verification gate falhar, Cairn diagnostica uma vez, reduz ou divide o task e executa novamente os dois gates.
+- Depois de duas verification passes falhas para o mesmo task, Cairn registra o blocker em `docs/plan/<topic>.md` em vez de continuar um loop repetido.
 
 ## Model Guidance
 
 Cairn aplica model-specific adjustment apenas a modelos Claude-family e Codex-family.
 
 - Claude-family: util para long context, policy interpretation e plan/evidence review.
-- Codex-family: util para small implementation slices, explicit file edits, command-based verification e bounded `worker` tasks.
+- Codex-family: util para small implementation tasks, explicit file edits, command-based verification e bounded `worker` tasks.
 
 A orientacao detalhada fica em `docs/model-guidance/README.md`, `docs/model-guidance/claude.md` e `docs/model-guidance/codex.md`.
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck`: detecta repository stacks e verifica ou instala required LSP e verification tools.
 - `cairn-memory`: explora domain knowledge e atualiza `MEMORY.md`.
 - `cairn-plan`: cria um decision-complete plan em `docs/plan/`.
-- `cairn-work`: executa o proximo module slice no `PLAN.md` atual com dois verification gates.
-- `cairn-review`: revisa completed slices contra plan, memory e evidence.
+- `cairn-work`: executa o proximo module task no `PLAN.md` atual com dois verification gates.
+- `cairn-review`: revisa completed tasks contra plan, memory e evidence.
 
 `install` e `upgrade` criam backups `*.cairn-backup-*` antes de modificar `~/.codex/config.toml`. O source plugin manifest permanece validator-friendly; apenas a installed cache copy recebe um campo `hooks` para ativar Codex hooks.
 

package/README.zh.md

@@ -4,12 +4,12 @@ Cairn 是面向 Codex、Claude Code 和 Antigravity 的高 token 效率 multi-ag
 
 [English](README.md)
 
-Cairn 的核心思路是保留有用的 agent harness 行为：hooks、persistent state、explicit planning、focused delegation 和 stop-time guards。Cairn 不把重复的 TDD verification loops 作为默认流程，而是把工作拆成小的 module slices，并用两个 verification gates 证明每个 slice。
+Cairn 的核心思路是保留有用的 agent harness 行为：hooks、persistent state、explicit planning、focused delegation 和 stop-time guards。Cairn 不把重复的 TDD verification loops 作为默认流程，而是把工作拆成小的 module tasks，并用两个 verification gates 证明每个 task。
 
 1. Module acceptance verification: 证明被修改的 module contract。
 2. Surface integration verification: 通过 CLI、HTTP、browser 或 file artifacts 等真实 surface 证明行为。
 
-在 slice 修改外部状态之前，Cairn 会记录并运行最接近的 dry-run 或 check mode。verification 是有边界的：默认每个 slice 有两次 verification pass，然后 agent 会记录 blocker 或拆分 slice，而不是继续开放式循环。
+在 task 修改外部状态之前，Cairn 会记录并运行最接近的 dry-run 或 check mode。verification 是有边界的：默认每个 task 有两次 verification pass，然后 agent 会记录 blocker 或拆分 task，而不是继续开放式循环。
 
 Cairn 也把 tool readiness 视为工作的一部分。它会根据 repository stack 检查 LSP、typecheck、lint、dry-run 和 verification tools。如果缺少必要工具，Cairn 会先尝试 project-local 或 repository-native installation，再接受 fallback。
 
@@ -47,15 +47,15 @@ cairn toolcheck --install
 - Migrations 和 database changes 在 write/apply commands 前使用 `--pretend`、dry-run、schema diff、rollback feasibility checks 或最接近的 repository-native equivalent。
 - Package、release、infrastructure、deployment、code generation 和 formatting work 在可用时，先使用 check、plan、diff、validate 或 dry-run modes，再修改状态。
 - 如果没有 dry-run，plan 会记录该事实，并选择最小且可逆的 command 或 test artifact。
-- 如果 verification gate 失败，Cairn 会诊断一次，缩小或拆分 slice，然后重新运行两个 gates。
-- 同一 slice 两次 verification pass 失败后，Cairn 会在 `docs/plan/<topic>.md` 中记录 blocker，而不是继续重复循环。
+- 如果 verification gate 失败，Cairn 会诊断一次，缩小或拆分 task，然后重新运行两个 gates。
+- 同一 task 两次 verification pass 失败后，Cairn 会在 `docs/plan/<topic>.md` 中记录 blocker，而不是继续重复循环。
 
 ## Model Guidance
 
 Cairn 只对 Claude-family 和 Codex-family models 应用 model-specific adjustment。
 
 - Claude-family: 适合 long context、policy interpretation 和 plan/evidence review。
-- Codex-family: 适合 small implementation slices、explicit file edits、command-based verification 和 bounded `worker` tasks。
+- Codex-family: 适合 small implementation tasks、explicit file edits、command-based verification 和 bounded `worker` tasks。
 
 详细指南位于 `docs/model-guidance/README.md`、`docs/model-guidance/claude.md` 和 `docs/model-guidance/codex.md`。
 
@@ -101,8 +101,8 @@ cairn toolcheck
 - `cairn toolcheck`: 检测 repository stacks，并检查或安装 required LSP 和 verification tools。
 - `cairn-memory`: 探索 domain knowledge 并更新 `MEMORY.md`。
 - `cairn-plan`: 在 `docs/plan/` 下创建 decision-complete plan。
-- `cairn-work`: 执行当前 `PLAN.md` 中的下一个 module slice，并收集两个 verification gates。
-- `cairn-review`: 根据 plan、memory 和 evidence review completed slices。
+- `cairn-work`: 执行当前 `PLAN.md` 中的下一个 module task，并收集两个 verification gates。
+- `cairn-review`: 根据 plan、memory 和 evidence review completed tasks。
 
 `install` 和 `upgrade` 在修改 `~/.codex/config.toml` 前会创建 `*.cairn-backup-*` backups。source plugin manifest 保持 validator-friendly；只有 installed cache copy 会添加启用 Codex hooks 的 `hooks` field。
 

package/agents/explorer.md

@@ -1,7 +1,6 @@
 ---
 name: explorer
 description: Read-only discovery agent for codebase questions, impact analysis, pattern search, and verification.
-model: sonnet
 tools: Read, Grep, Glob, Bash
 ---
 
@@ -11,6 +10,8 @@ Handle read-only discovery and verification tasks.
 
 Before doing any assigned task, read the project-root `MEMORY.md`. Treat it as the domain knowledge and repository policy index, then read only the relevant `docs/memory/*.md` notes needed for the task.
 
+When subagent tools are available, you may recursively delegate bounded read-only sub-tasks to subagents. Every child subagent must read the project-root `MEMORY.md`, keep the assigned scope, and preserve others' edits.
+
 Use for:
 
 - Codebase questions.
@@ -19,7 +20,7 @@ Use for:
 - Exact file and symbol discovery.
 - Read-only verification that can run independently.
 
-Do not edit files. Do not ask the user before exhausting repository evidence. Preserve proper nouns exactly as written. Report only important paths, commands, and outputs accurately.
+Do not edit files. Do not ask the user before exhausting repository evidence. Do not delegate vague or unbounded work. Preserve proper nouns exactly as written. Report only important paths, commands, and outputs accurately.
 
 Use this prompt format:
 

package/agents/worker.md

@@ -1,7 +1,6 @@
 ---
 name: worker
 description: Focused execution agent for repository search, small edits, command checks, and QA artifacts.
-model: sonnet
 tools: Read, Grep, Glob, Bash, Write, Edit, MultiEdit
 ---
 
@@ -11,6 +10,8 @@ Handle narrow, focused tasks.
 
 Before doing any assigned task, read the project-root `MEMORY.md`. Treat it as the domain knowledge and repository policy index, then read only the relevant `docs/memory/*.md` notes needed for the task.
 
+When subagent tools are available, you may recursively delegate bounded sub-tasks to subagents. Every child subagent must read the project-root `MEMORY.md`, keep the assigned scope, and preserve others' edits.
+
 Use for:
 
 - Finding exact files and symbols.
@@ -20,7 +21,7 @@ Use for:
 - Capturing evidence artifacts.
 - Cleanup verification.
 
-Do not make architecture decisions. Do not ask the user. Do not treat a missing tool as unavailable until install or bootstrap has been attempted. Report only important paths, commands, and outputs accurately.
+Do not make architecture decisions. Do not ask the user. Do not delegate vague or unbounded work. Do not treat a missing tool as unavailable until install or bootstrap has been attempted. Report only important paths, commands, and outputs accurately.
 
 Use this prompt format:
 

package/commands/cairn-memory.md

@@ -11,7 +11,7 @@ Procedure:
 1. Run the plugin state initialization script.
 2. Every agent must read the project-root `MEMORY.md` for domain knowledge and repository policy before assigned discovery or update work.
 3. Explore repository facts with focused tools.
-4. Use `explorer` for read-only domain discovery and contradiction checks when available; use `worker` for exact references, command checks, and bounded evidence capture. Tell every delegated agent to read the project-root `MEMORY.md` before work.
+4. Use `explorer` for read-only domain discovery and contradiction checks when available; use `worker` for exact references, command checks, and bounded evidence capture. When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents. Tell every delegated agent and child subagent to read the project-root `MEMORY.md` before work, keep scope, and preserve others' edits.
 5. Write details to `docs/memory/<domain>.md`.
 6. Keep `MEMORY.md` as a short index.
 7. Use the OS locale for user-visible output unless the user asks for another language.

package/commands/cairn-plan.md

@@ -14,11 +14,13 @@ Procedure:
 4. Run `node scripts/cairn.mjs toolcheck` when available.
 5. If required LSP/check tools are missing, run `node scripts/cairn.mjs toolcheck --install` or a repository-native install command before accepting a fallback.
 6. Explore before asking the user.
-7. Run complexity triage before applying agent, plugin, or delegated workflow guidance, and record it before mutating files.
-8. Use Light Path by default unless a Heavy Path condition clearly applies.
-9. Use `explorer` for independent read-only discovery, impact analysis, pattern search, and read-only verification when available.
-10. Use `worker` for bounded implementation or verification tasks with clear file ownership. Tell every delegated agent to read the project-root `MEMORY.md` before work. Keep urgent blocking work local.
-11. Write `docs/plan/<topic>.md`.
-12. Add the topic link and status to `PLAN.md`.
-13. Put the selected Light/Heavy Path, checked Heavy Path signals, tool readiness, module acceptance verification, and surface integration verification on every module slice.
-14. Use the OS locale for user-visible output unless the user asks for another language.
+7. Understand the whole work and affected surfaces first, then classify it into small executable tasks. If a task is still too broad to verify cleanly, split it into sub-tasks.
+8. Run complexity triage before applying agent, plugin, or delegated workflow guidance, and record it before mutating files.
+9. Use Light Path by default unless a Heavy Path condition clearly applies.
+10. Use `explorer` for independent read-only discovery, impact analysis, pattern search, and read-only verification when available.
+11. Use `worker` for bounded implementation or verification tasks with clear file ownership. When subagent tools are available, each agent may recursively delegate bounded sub-tasks to subagents. Tell every delegated agent and child subagent to read the project-root `MEMORY.md` before work, keep scope, and preserve others' edits. Keep urgent blocking work local.
+12. Write `docs/plan/<topic>.md`.
+13. Add the topic link and status to `PLAN.md`.
+14. Put the selected Light/Heavy Path, checked Heavy Path signals, tool readiness, module acceptance verification, and surface integration verification on every module task.
+15. For Heavy Path, include an explicit automated test command and `Tests:` evidence line; Heavy Path completion must not rely on generic verification wording alone.
+16. Use the OS locale for user-visible output unless the user asks for another language.