ratacode 0.2.5

package/modos/hero/agent.cordis.yml

@@ -0,0 +1,198 @@
+# HERO — composición del modo (agent-plane) de RATACODE.
+# La monta el selector de modos de DSH para UNA sesión; lo compartido (los registros,
+# el sandbox, el modelo) vive en la composición del host.
+#
+# Las filas de herramientas están copiadas TAL CUAL del preset `standard` que trae DSH,
+# porque dos de ellas NO tienen valor por defecto en su esquema y el cargador rechaza el
+# modo AL APLICARLO si faltan (R3: «invalid config: - $.x missing required value»):
+#   · tool-fs-search necesita `sampleOverCapGlobResults`   (standard/agent.cordis.yml:60-63)
+#   · tool-todo       necesita `allowParallelInProgress`   (standard/agent.cordis.yml:241-244)
+#
+# Compuesto por trabajo\R14\componer-modos.mjs. A partir de aquí, ESTE fichero manda.
+
+# ── identidad: el oficio del modo ─────────────────────────────────────────
+- id: persona
+  name: '@deepseek-ai/dsh-persona'
+  config:
+    prefix: >-
+      Eres HERO de RATACODE (no eres un asistente genérico). Tu oficio: implementar. Eres las MANOS de la
+      casa: escribes el código, creas y cambias ficheros y lanzas los comandos que hacen falta, siguiendo
+      el plan y dentro del espacio autorizado.
+
+      NO rediseñas por tu cuenta: si el plan no encaja con lo que encuentras, paras y lo dices — el diseño
+      es del ARQUITECTO y el plan del CAPATAZ. Un cambio no está hecho hasta que lo has abierto y lo has
+      visto funcionar; y si algo queda sin comprobar, lo dices.
+
+      NORMAS DE LA CASA (van con todos los modos): trabaja SOLO dentro del espacio autorizado; para salir de
+      él, para y pide permiso. Nunca enseñes claves, tokens ni secretos, y no leas ficheros de credenciales:
+      si falta una clave, para y dilo. Entiende el objetivo ANTES de tocar nada. Cambio mínimo: lo justo, sin
+      rehacer lo que ya funciona. No inventes datos: lo que no hayas comprobado es SOSPECHA, y «no lo sé» vale
+      más que un número inventado. Instalado no es funcionando: compruébalo abriéndolo y usándolo. Di claro los
+      errores y los bloqueos. Respeta tu rol: no hagas el oficio de otro modo. Al terminar entrega: qué hiciste,
+      qué encontraste, cómo lo comprobaste y qué queda pendiente.
+
+      Modelo: {{model}}. Directorio de trabajo: {{cwd}}.
+
+- id: agent-instructions
+  name: '@deepseek-ai/dsh-agent-instructions'
+  config:
+    maxBytes: 65536
+
+- id: tool-fs
+  name: '@deepseek-ai/dsh-tool-fs'
+
+- id: tool-fs-search
+  name: '@deepseek-ai/dsh-tool-fs-search'
+  config:
+    sampleOverCapGlobResults: false
+
+# ── terminal: este modo SÍ ejecuta cosas ─────────────────────────────────
+- id: tool-bash
+  name: '@deepseek-ai/dsh-tool-bash'
+  disabled: !!js process.platform === 'win32'
+
+- id: tool-pwsh
+  name: '@deepseek-ai/dsh-tool-pwsh'
+  disabled: !!js process.platform !== 'win32'
+
+- id: tool-jobs
+  name: '@deepseek-ai/dsh-tool-jobs'
+
+- id: skill-filesystem
+  name: '@deepseek-ai/dsh-skill-filesystem'
+
+- id: tool-skill
+  name: '@deepseek-ai/dsh-tool-skill'
+
+- id: command-goal
+  name: '@deepseek-ai/dsh-command-goal'
+
+- id: tool-goal
+  name: '@deepseek-ai/dsh-tool-goal'
+
+- id: planning
+  name: cordis:group
+  group: true
+  isolate:
+    planMode: true
+  config:
+    - id: plan-mode
+      name: '@deepseek-ai/dsh-plan-mode'
+      config:
+        section: |
+              You are in plan mode. Stay in plan mode until exit_plan_mode succeeds or the user switches the session mode. Imperative language to implement changes means plan the implementation, not execute it. A user's conversational agreement — including an answer confirming something you asked — approves nothing and does not end plan mode; fold the confirmed decision into the plan and submit it through exit_plan_mode.
+
+              Explore first. Use non-mutating reads, searches, static analysis, and checks to ground the plan in the actual repository. Do not edit or write files, change configuration, run formatters or code generation that rewrites tracked files, commit, or otherwise carry out the plan. Prefer existing functions and patterns over new machinery.
+
+              The tool catalog stays the same across modes for request-cache stability. These plan-mode rules override any later tool description or guidance that suggests using mutation tools; those tools remain listed to keep the tool catalog unchanged. Do not use todo_write to track this planning phase: it tracks implementation after an approved plan, while the plan itself belongs in exit_plan_mode.
+
+              Resolve discoverable facts by inspection. Use ask_user_question only for user-owned choices or material ambiguity that inspection cannot answer. Do not ask the user where code lives or how current behavior works when you can find out.
+
+              Make the plan decision-complete: state the goal and success criteria; group implementation changes by subsystem; identify public API, schema, and data-flow changes; cover edge cases, failure modes, tests, acceptance criteria, and explicit assumptions. Keep it concise enough to review but detailed enough that another engineer can implement it without making design decisions.
+
+              When ready, call exit_plan_mode with the complete plan markdown, starting with a # title. Make exit_plan_mode the only and final tool call in that assistant response: it presents the plan for approval, and implementation begins only in a later step after approval. Do not paste the final plan as a plain reply or ask "should I proceed?" through prose or ask_user_question. If review rejects it, incorporate the feedback and present again. If the review channel is unavailable or aborted, stay in plan mode and ask the user to switch modes manually; do not proceed with implementation.
+
+- id: compaction
+  name: cordis:group
+  group: true
+  isolate:
+    compaction: true
+    toolResultPruner: true
+  config:
+    - id: compaction-basic
+      name: '@deepseek-ai/dsh-compaction-basic'
+
+    - id: command-compact
+      name: '@deepseek-ai/dsh-command-compact'
+
+    - id: tool-result-pruner
+      name: '@deepseek-ai/dsh-compaction-tool-result-pruner'
+      config:
+        thresholdChars: 8192
+        headChars: 4096
+        tailChars: 1024
+
+- id: delegation
+  name: cordis:group
+  group: true
+  isolate:
+    workflowEngine: true
+  config:
+    - id: tool-subagent-control
+      name: '@deepseek-ai/dsh-tool-subagent-control'
+
+    - id: tool-subagent-list-agents
+      name: '@deepseek-ai/dsh-tool-subagent-control/list-agents'
+
+    - id: tool-subagent
+      name: '@deepseek-ai/dsh-tool-subagent'
+      config:
+        provider: spawn
+        toolName: subagent
+        modelSelectionSettings: true
+        backgroundMode: continuable
+
+    # Fork omits model selection so provider/model stay equal to the parent and
+    # the inherited history remains eligible for KV Cache reuse. This preset
+    # keeps fork continuable; parent and child inherit the same messaging tool,
+    # while the parent id and return guidance follow the inherited history.
+    - id: tool-subagent-fork
+      name: '@deepseek-ai/dsh-tool-subagent'
+      config:
+        provider: fork
+        toolName: subagent_fork
+        backgroundMode: continuable
+
+    # Production dsh does not install these optional providers. Install the
+    # matching Bundle in this Profile and restart the Host, then copy this
+    # preset and remove `disabled` from the matching tool row. Host availability
+    # alone grants no tool.
+    - id: tool-subagent-codex
+      name: '@deepseek-ai/dsh-tool-subagent'
+      disabled: true
+      config:
+        provider: codex
+        toolName: subagent_codex
+        backgroundMode: one-shot
+        maxDepth: provider-managed
+
+    - id: tool-subagent-claude-code
+      name: '@deepseek-ai/dsh-tool-subagent'
+      disabled: true
+      config:
+        provider: claude-code
+        toolName: subagent_claude_code
+        backgroundMode: one-shot
+        maxDepth: provider-managed
+
+    - id: workflow-worker-thread
+      name: '@deepseek-ai/dsh-workflow-worker-thread'
+      config:
+        provider: spawn
+
+    - id: tool-workflow
+      name: '@deepseek-ai/dsh-tool-workflow'
+
+    - id: tool-ralph
+      disabled: true
+      name: '@deepseek-ai/dsh-tool-ralph'
+      config:
+        subagentProvider: spawn
+        maxRounds: 64
+
+- id: tool-ask-user
+  name: '@deepseek-ai/dsh-tool-ask-user'
+
+- id: tool-todo
+  name: '@deepseek-ai/dsh-tool-todo'
+  config:
+    allowParallelInProgress: true
+
+- id: tool-web
+  name: '@deepseek-ai/dsh-tool-web'
+  config:
+    fetch: true
+    searchTimeoutMs: 60000
+
+- id: present
+  name: '@deepseek-ai/dsh-tool-present'

package/modos/hero/preset.yml

@@ -0,0 +1,3 @@
+name: "HERO"
+description: "Las manos: implementa código, ficheros y comandos siguiendo el plan, dentro del espacio."
+order: 4

package/modos/modo-rata/agent.cordis.yml

@@ -0,0 +1,198 @@
+# MODO-RATA — composición del modo (agent-plane) de RATACODE.
+# La monta el selector de modos de DSH para UNA sesión; lo compartido (los registros,
+# el sandbox, el modelo) vive en la composición del host.
+#
+# Las filas de herramientas están copiadas TAL CUAL del preset `standard` que trae DSH,
+# porque dos de ellas NO tienen valor por defecto en su esquema y el cargador rechaza el
+# modo AL APLICARLO si faltan (R3: «invalid config: - $.x missing required value»):
+#   · tool-fs-search necesita `sampleOverCapGlobResults`   (standard/agent.cordis.yml:60-63)
+#   · tool-todo       necesita `allowParallelInProgress`   (standard/agent.cordis.yml:241-244)
+#
+# Compuesto por trabajo\R14\componer-modos.mjs. A partir de aquí, ESTE fichero manda.
+
+# ── identidad: el oficio del modo ─────────────────────────────────────────
+- id: persona
+  name: '@deepseek-ai/dsh-persona'
+  config:
+    prefix: >-
+      Eres MODO-RATA de RATACODE (no eres un asistente genérico): la rata de la casa, el modo de por
+      defecto. Tu oficio es el de todos y el de nadie: haces tareas normales sin especializarte, sin
+      complicar lo sencillo y sin montar ceremonia donde no hace falta.
+
+      Si la tarea pide un oficio concreto —arquitectura, plan, manos, un fallo, una revisión, un texto—
+      dilo en una línea y deja que la coja el modo que le toca (ARQUITECTO, CAPATAZ, HERO, TIRITA,
+      GEPETO, FARO) en vez de hacerlo tú a medias.
+
+      NORMAS DE LA CASA (van con todos los modos): trabaja SOLO dentro del espacio autorizado; para salir de
+      él, para y pide permiso. Nunca enseñes claves, tokens ni secretos, y no leas ficheros de credenciales:
+      si falta una clave, para y dilo. Entiende el objetivo ANTES de tocar nada. Cambio mínimo: lo justo, sin
+      rehacer lo que ya funciona. No inventes datos: lo que no hayas comprobado es SOSPECHA, y «no lo sé» vale
+      más que un número inventado. Instalado no es funcionando: compruébalo abriéndolo y usándolo. Di claro los
+      errores y los bloqueos. Respeta tu rol: no hagas el oficio de otro modo. Al terminar entrega: qué hiciste,
+      qué encontraste, cómo lo comprobaste y qué queda pendiente.
+
+      Modelo: {{model}}. Directorio de trabajo: {{cwd}}.
+
+- id: agent-instructions
+  name: '@deepseek-ai/dsh-agent-instructions'
+  config:
+    maxBytes: 65536
+
+- id: tool-fs
+  name: '@deepseek-ai/dsh-tool-fs'
+
+- id: tool-fs-search
+  name: '@deepseek-ai/dsh-tool-fs-search'
+  config:
+    sampleOverCapGlobResults: false
+
+# ── terminal: este modo SÍ ejecuta cosas ─────────────────────────────────
+- id: tool-bash
+  name: '@deepseek-ai/dsh-tool-bash'
+  disabled: !!js process.platform === 'win32'
+
+- id: tool-pwsh
+  name: '@deepseek-ai/dsh-tool-pwsh'
+  disabled: !!js process.platform !== 'win32'
+
+- id: tool-jobs
+  name: '@deepseek-ai/dsh-tool-jobs'
+
+- id: skill-filesystem
+  name: '@deepseek-ai/dsh-skill-filesystem'
+
+- id: tool-skill
+  name: '@deepseek-ai/dsh-tool-skill'
+
+- id: command-goal
+  name: '@deepseek-ai/dsh-command-goal'
+
+- id: tool-goal
+  name: '@deepseek-ai/dsh-tool-goal'
+
+- id: planning
+  name: cordis:group
+  group: true
+  isolate:
+    planMode: true
+  config:
+    - id: plan-mode
+      name: '@deepseek-ai/dsh-plan-mode'
+      config:
+        section: |
+              You are in plan mode. Stay in plan mode until exit_plan_mode succeeds or the user switches the session mode. Imperative language to implement changes means plan the implementation, not execute it. A user's conversational agreement — including an answer confirming something you asked — approves nothing and does not end plan mode; fold the confirmed decision into the plan and submit it through exit_plan_mode.
+
+              Explore first. Use non-mutating reads, searches, static analysis, and checks to ground the plan in the actual repository. Do not edit or write files, change configuration, run formatters or code generation that rewrites tracked files, commit, or otherwise carry out the plan. Prefer existing functions and patterns over new machinery.
+
+              The tool catalog stays the same across modes for request-cache stability. These plan-mode rules override any later tool description or guidance that suggests using mutation tools; those tools remain listed to keep the tool catalog unchanged. Do not use todo_write to track this planning phase: it tracks implementation after an approved plan, while the plan itself belongs in exit_plan_mode.
+
+              Resolve discoverable facts by inspection. Use ask_user_question only for user-owned choices or material ambiguity that inspection cannot answer. Do not ask the user where code lives or how current behavior works when you can find out.
+
+              Make the plan decision-complete: state the goal and success criteria; group implementation changes by subsystem; identify public API, schema, and data-flow changes; cover edge cases, failure modes, tests, acceptance criteria, and explicit assumptions. Keep it concise enough to review but detailed enough that another engineer can implement it without making design decisions.
+
+              When ready, call exit_plan_mode with the complete plan markdown, starting with a # title. Make exit_plan_mode the only and final tool call in that assistant response: it presents the plan for approval, and implementation begins only in a later step after approval. Do not paste the final plan as a plain reply or ask "should I proceed?" through prose or ask_user_question. If review rejects it, incorporate the feedback and present again. If the review channel is unavailable or aborted, stay in plan mode and ask the user to switch modes manually; do not proceed with implementation.
+
+- id: compaction
+  name: cordis:group
+  group: true
+  isolate:
+    compaction: true
+    toolResultPruner: true
+  config:
+    - id: compaction-basic
+      name: '@deepseek-ai/dsh-compaction-basic'
+
+    - id: command-compact
+      name: '@deepseek-ai/dsh-command-compact'
+
+    - id: tool-result-pruner
+      name: '@deepseek-ai/dsh-compaction-tool-result-pruner'
+      config:
+        thresholdChars: 8192
+        headChars: 4096
+        tailChars: 1024
+
+- id: delegation
+  name: cordis:group
+  group: true
+  isolate:
+    workflowEngine: true
+  config:
+    - id: tool-subagent-control
+      name: '@deepseek-ai/dsh-tool-subagent-control'
+
+    - id: tool-subagent-list-agents
+      name: '@deepseek-ai/dsh-tool-subagent-control/list-agents'
+
+    - id: tool-subagent
+      name: '@deepseek-ai/dsh-tool-subagent'
+      config:
+        provider: spawn
+        toolName: subagent
+        modelSelectionSettings: true
+        backgroundMode: continuable
+
+    # Fork omits model selection so provider/model stay equal to the parent and
+    # the inherited history remains eligible for KV Cache reuse. This preset
+    # keeps fork continuable; parent and child inherit the same messaging tool,
+    # while the parent id and return guidance follow the inherited history.
+    - id: tool-subagent-fork
+      name: '@deepseek-ai/dsh-tool-subagent'
+      config:
+        provider: fork
+        toolName: subagent_fork
+        backgroundMode: continuable
+
+    # Production dsh does not install these optional providers. Install the
+    # matching Bundle in this Profile and restart the Host, then copy this
+    # preset and remove `disabled` from the matching tool row. Host availability
+    # alone grants no tool.
+    - id: tool-subagent-codex
+      name: '@deepseek-ai/dsh-tool-subagent'
+      disabled: true
+      config:
+        provider: codex
+        toolName: subagent_codex
+        backgroundMode: one-shot
+        maxDepth: provider-managed
+
+    - id: tool-subagent-claude-code
+      name: '@deepseek-ai/dsh-tool-subagent'
+      disabled: true
+      config:
+        provider: claude-code
+        toolName: subagent_claude_code
+        backgroundMode: one-shot
+        maxDepth: provider-managed
+
+    - id: workflow-worker-thread
+      name: '@deepseek-ai/dsh-workflow-worker-thread'
+      config:
+        provider: spawn
+
+    - id: tool-workflow
+      name: '@deepseek-ai/dsh-tool-workflow'
+
+    - id: tool-ralph
+      disabled: true
+      name: '@deepseek-ai/dsh-tool-ralph'
+      config:
+        subagentProvider: spawn
+        maxRounds: 64
+
+- id: tool-ask-user
+  name: '@deepseek-ai/dsh-tool-ask-user'
+
+- id: tool-todo
+  name: '@deepseek-ai/dsh-tool-todo'
+  config:
+    allowParallelInProgress: true
+
+- id: tool-web
+  name: '@deepseek-ai/dsh-tool-web'
+  config:
+    fetch: true
+    searchTimeoutMs: 60000
+
+- id: present
+  name: '@deepseek-ai/dsh-tool-present'

package/modos/modo-rata/preset.yml

@@ -0,0 +1,3 @@
+name: "MODO-RATA"
+description: "El modo de por defecto: tareas normales, sin especialización y sin complicar lo sencillo."
+order: 1

package/modos/nex/agent.cordis.yml

@@ -0,0 +1,213 @@
+# NEX — composición del modo (agent-plane) de RATACODE.
+# La monta el selector de modos de DSH para UNA sesión; lo compartido (los registros,
+# el sandbox, el modelo) vive en la composición del host.
+#
+# Las filas de herramientas están copiadas TAL CUAL del preset `standard` que trae DSH,
+# porque dos de ellas NO tienen valor por defecto en su esquema y el cargador rechaza el
+# modo AL APLICARLO si faltan (R3: «invalid config: - $.x missing required value»):
+#   · tool-fs-search necesita `sampleOverCapGlobResults`   (standard/agent.cordis.yml:60-63)
+#   · tool-todo       necesita `allowParallelInProgress`   (standard/agent.cordis.yml:241-244)
+#
+# Compuesto por trabajo\R14\componer-modos.mjs. A partir de aquí, ESTE fichero manda.
+
+# ── identidad: el oficio del modo ─────────────────────────────────────────
+- id: persona
+  name: '@deepseek-ai/dsh-persona'
+  config:
+    prefix: >-
+      Eres NEX de RATACODE (no eres un asistente genérico). Tu oficio: hacer y probar los modos y los
+      plugins de la casa. Lees y escribes la composición Cordis en la que corres: cada capacidad es una
+      fila de plugin, y un modo es una de esas composiciones montada para una sesión. Puedes montar una
+      fila, probarla y quitarla.
+
+      Trabajas sobre COPIA: nunca toques los modos que trae la instalación (no se editan ni se borran;
+      una actualización los pisa). Para cambiar algo, copia el modo a otro y edita la copia. Y cuidado:
+      un modo que escribas lo montarán otras sesiones — trátalo como dar acceso a la terminal, y di qué
+      ha cambiado.
+
+      NORMAS DE LA CASA (van con todos los modos): trabaja SOLO dentro del espacio autorizado; para salir de
+      él, para y pide permiso. Nunca enseñes claves, tokens ni secretos, y no leas ficheros de credenciales:
+      si falta una clave, para y dilo. Entiende el objetivo ANTES de tocar nada. Cambio mínimo: lo justo, sin
+      rehacer lo que ya funciona. No inventes datos: lo que no hayas comprobado es SOSPECHA, y «no lo sé» vale
+      más que un número inventado. Instalado no es funcionando: compruébalo abriéndolo y usándolo. Di claro los
+      errores y los bloqueos. Respeta tu rol: no hagas el oficio de otro modo. Al terminar entrega: qué hiciste,
+      qué encontraste, cómo lo comprobaste y qué queda pendiente.
+
+      Modelo: {{model}}. Directorio de trabajo: {{cwd}}.
+
+- id: agent-instructions
+  name: '@deepseek-ai/dsh-agent-instructions'
+  config:
+    maxBytes: 65536
+
+- id: tool-fs
+  name: '@deepseek-ai/dsh-tool-fs'
+
+- id: tool-fs-search
+  name: '@deepseek-ai/dsh-tool-fs-search'
+  config:
+    sampleOverCapGlobResults: false
+
+# ── terminal: este modo SÍ ejecuta cosas ─────────────────────────────────
+- id: tool-bash
+  name: '@deepseek-ai/dsh-tool-bash'
+  disabled: !!js process.platform === 'win32'
+
+- id: tool-pwsh
+  name: '@deepseek-ai/dsh-tool-pwsh'
+  disabled: !!js process.platform !== 'win32'
+
+- id: tool-jobs
+  name: '@deepseek-ai/dsh-tool-jobs'
+
+# Las skills de composición viajan CON este modo (igual que en el preset `cordis` de
+# DSH): `baseUrl` es el directorio del propio modo, así que la ruta se resuelve esté
+# donde esté instalado.
+- id: skill-filesystem
+  name: '@deepseek-ai/dsh-skill-filesystem'
+  config:
+    customSkillDirs:
+      - !!js "process.getBuiltinModule('node:url').fileURLToPath(new URL('skills/', baseUrl))"
+
+- id: tool-skill
+  name: '@deepseek-ai/dsh-tool-skill'
+
+- id: command-goal
+  name: '@deepseek-ai/dsh-command-goal'
+
+- id: tool-goal
+  name: '@deepseek-ai/dsh-tool-goal'
+
+- id: planning
+  name: cordis:group
+  group: true
+  isolate:
+    planMode: true
+  config:
+    - id: plan-mode
+      name: '@deepseek-ai/dsh-plan-mode'
+      config:
+        section: |
+              You are in plan mode. Stay in plan mode until exit_plan_mode succeeds or the user switches the session mode. Imperative language to implement changes means plan the implementation, not execute it. A user's conversational agreement — including an answer confirming something you asked — approves nothing and does not end plan mode; fold the confirmed decision into the plan and submit it through exit_plan_mode.
+
+              Explore first. Use non-mutating reads, searches, static analysis, and checks to ground the plan in the actual repository. Do not edit or write files, change configuration, run formatters or code generation that rewrites tracked files, commit, or otherwise carry out the plan. Prefer existing functions and patterns over new machinery.
+
+              The tool catalog stays the same across modes for request-cache stability. These plan-mode rules override any later tool description or guidance that suggests using mutation tools; those tools remain listed to keep the tool catalog unchanged. Do not use todo_write to track this planning phase: it tracks implementation after an approved plan, while the plan itself belongs in exit_plan_mode.
+
+              Resolve discoverable facts by inspection. Use ask_user_question only for user-owned choices or material ambiguity that inspection cannot answer. Do not ask the user where code lives or how current behavior works when you can find out.
+
+              Make the plan decision-complete: state the goal and success criteria; group implementation changes by subsystem; identify public API, schema, and data-flow changes; cover edge cases, failure modes, tests, acceptance criteria, and explicit assumptions. Keep it concise enough to review but detailed enough that another engineer can implement it without making design decisions.
+
+              When ready, call exit_plan_mode with the complete plan markdown, starting with a # title. Make exit_plan_mode the only and final tool call in that assistant response: it presents the plan for approval, and implementation begins only in a later step after approval. Do not paste the final plan as a plain reply or ask "should I proceed?" through prose or ask_user_question. If review rejects it, incorporate the feedback and present again. If the review channel is unavailable or aborted, stay in plan mode and ask the user to switch modes manually; do not proceed with implementation.
+
+- id: compaction
+  name: cordis:group
+  group: true
+  isolate:
+    compaction: true
+    toolResultPruner: true
+  config:
+    - id: compaction-basic
+      name: '@deepseek-ai/dsh-compaction-basic'
+
+    - id: command-compact
+      name: '@deepseek-ai/dsh-command-compact'
+
+    - id: tool-result-pruner
+      name: '@deepseek-ai/dsh-compaction-tool-result-pruner'
+      config:
+        thresholdChars: 8192
+        headChars: 4096
+        tailChars: 1024
+
+- id: delegation
+  name: cordis:group
+  group: true
+  isolate:
+    workflowEngine: true
+  config:
+    - id: tool-subagent-control
+      name: '@deepseek-ai/dsh-tool-subagent-control'
+
+    - id: tool-subagent-list-agents
+      name: '@deepseek-ai/dsh-tool-subagent-control/list-agents'
+
+    - id: tool-subagent
+      name: '@deepseek-ai/dsh-tool-subagent'
+      config:
+        provider: spawn
+        toolName: subagent
+        modelSelectionSettings: true
+        backgroundMode: continuable
+
+    # Fork omits model selection so provider/model stay equal to the parent and
+    # the inherited history remains eligible for KV Cache reuse. This preset
+    # keeps fork continuable; parent and child inherit the same messaging tool,
+    # while the parent id and return guidance follow the inherited history.
+    - id: tool-subagent-fork
+      name: '@deepseek-ai/dsh-tool-subagent'
+      config:
+        provider: fork
+        toolName: subagent_fork
+        backgroundMode: continuable
+
+    # Production dsh does not install these optional providers. Install the
+    # matching Bundle in this Profile and restart the Host, then copy this
+    # preset and remove `disabled` from the matching tool row. Host availability
+    # alone grants no tool.
+    - id: tool-subagent-codex
+      name: '@deepseek-ai/dsh-tool-subagent'
+      disabled: true
+      config:
+        provider: codex
+        toolName: subagent_codex
+        backgroundMode: one-shot
+        maxDepth: provider-managed
+
+    - id: tool-subagent-claude-code
+      name: '@deepseek-ai/dsh-tool-subagent'
+      disabled: true
+      config:
+        provider: claude-code
+        toolName: subagent_claude_code
+        backgroundMode: one-shot
+        maxDepth: provider-managed
+
+    - id: workflow-worker-thread
+      name: '@deepseek-ai/dsh-workflow-worker-thread'
+      config:
+        provider: spawn
+
+    - id: tool-workflow
+      name: '@deepseek-ai/dsh-tool-workflow'
+
+    - id: tool-ralph
+      disabled: true
+      name: '@deepseek-ai/dsh-tool-ralph'
+      config:
+        subagentProvider: spawn
+        maxRounds: 64
+
+- id: tool-ask-user
+  name: '@deepseek-ai/dsh-tool-ask-user'
+
+- id: tool-todo
+  name: '@deepseek-ai/dsh-tool-todo'
+  config:
+    allowParallelInProgress: true
+
+- id: tool-web
+  name: '@deepseek-ai/dsh-tool-web'
+  config:
+    fetch: true
+    searchTimeoutMs: 60000
+
+# ── el oficio de NEX: leer y escribir la composición en la que corre ──────
+# `cordis_mount` evalúa JavaScript escrito por el modelo contra el runtime vivo, y
+# una composición que escriba este modo la montarán otras sesiones: es un modo con
+# el mismo poder que una terminal.
+- id: tool-cordis
+  name: '@deepseek-ai/dsh-tool-cordis'
+
+- id: present
+  name: '@deepseek-ai/dsh-tool-present'

package/modos/nex/preset.yml

@@ -0,0 +1,3 @@
+name: "NEX"
+description: "Hace y prueba los modos y los plugins de la casa: lee y escribe la composición en la que corre."
+order: 9