@agentprojectcontext/apx 1.32.0 → 1.32.2

package/src/interfaces/web/src/i18n/en.ts

@@ -286,8 +286,8 @@ export const en = {
     chat: {
       title:       "Chat with agent",
       subtitle:    "Direct conversations with project agents. The super-agent does not intervene.",
-      roby_title:  "Chat with Roby",
-      roby_subtitle: "Chat with Roby — the APX super-agent. Can use tools (projects, tasks, mcps, agents).",
+      superagent_title:  "Chat with {persona}",
+      superagent_subtitle: "Chat with {persona} — the APX super-agent. Can use tools (projects, tasks, mcps, agents).",
       empty:       "Send a message to start the conversation.",
       placeholder: "Type something and press enter to send (shift+enter = new line)",
       send:        "Send",
@@ -710,13 +710,13 @@ export const en = {
     delete_btn:  "Delete",
   },
 
-  roby: {
-    title:          "Roby",
+  superagent: {
+    title:          "{persona}",
     badge:          "super-agent · APX",
     desc:           "Quick chat with your super-agent. Has access to tools (projects, tasks, mcps, agents); for a longer persistent thread, open Chats.",
-    empty:          "Send Roby a message to get started.",
-    thinking:       "Roby is thinking…",
-    talk:           "Talk to Roby",
+    empty:          "Send {persona} a message to get started.",
+    thinking:       "{persona} is thinking…",
+    talk:           "Talk to {persona}",
     new_chat:       "New chat",
     placeholder:    "Type and press enter to send (shift+enter = new line)…",
   },

package/src/interfaces/web/src/i18n/es.ts

@@ -287,8 +287,8 @@ export const es = {
     chat: {
       title:       "Chat con agente",
       subtitle:    "Conversaciones directas con agentes del proyecto. El super-agent no interviene.",
-      roby_title:  "Chat con Roby",
-      roby_subtitle: "Chat con Roby — el super-agente APX. Puede usar tools (proyectos, tasks, mcps, agentes).",
+      superagent_title:  "Chat con {persona}",
+      superagent_subtitle: "Chat con {persona} — el super-agente APX. Puede usar tools (proyectos, tasks, mcps, agentes).",
       empty:       "Mandá un mensaje para arrancar la conversación.",
       placeholder: "Escribí algo y enter para enviar (shift+enter = nueva línea)",
       send:        "Enviar",
@@ -711,13 +711,13 @@ export const es = {
     delete_btn:  "Borrar",
   },
 
-  roby: {
-    title:          "Roby",
+  superagent: {
+    title:          "{persona}",
     badge:          "super-agent · APX",
     desc:           "Conversación rápida con tu super-agente. Tiene acceso a tools (proyectos, tasks, mcps, agentes); para un hilo más largo y persistente, abrí Chats.",
-    empty:          "Mandale un mensaje a Roby para arrancar.",
-    thinking:       "Roby está pensando…",
-    talk:           "Hablar con Roby",
+    empty:          "Mandale un mensaje a {persona} para arrancar.",
+    thinking:       "{persona} está pensando…",
+    talk:           "Hablar con {persona}",
     new_chat:       "Nuevo chat",
     placeholder:    "Escribí y enter para enviar (shift+enter = nueva línea)…",
   },

package/src/interfaces/web/src/lib/api/skills.ts

@@ -0,0 +1,25 @@
+import { http } from "../http";
+
+export type SkillEntry = {
+  slug: string;
+  source: "bundled" | "user" | "project" | string;
+  description: string;
+};
+
+export type SkillsList = {
+  count: number;
+  skills: SkillEntry[];
+};
+
+export const Skills = {
+  /**
+   * List installed skills (bundled + user + optional project-scoped). The
+   * description is already condensed for one-line rendering.
+   */
+  list: (projectPath?: string) =>
+    http.get<SkillsList>(
+      projectPath
+        ? `/skills?project_path=${encodeURIComponent(projectPath)}`
+        : "/skills",
+    ),
+};

package/src/interfaces/web/src/lib/api.ts

@@ -22,6 +22,7 @@ export * from "./api/voice";
 export * from "./api/deck";
 export * from "./api/code";
 export * from "./api/artifacts";
+export * from "./api/skills";
 
 // Re-export the daemon types so older imports of "../lib/api" still work.
 export type {

package/src/interfaces/web/src/screens/modules/CodeScreen.tsx

@@ -1,8 +1,8 @@
 import { useCallback, useEffect, useMemo, useRef, useState } from "react";
 import useSWR from "swr";
-import { Code2 } from "lucide-react";
 import { Code, Projects } from "../../lib/api";
 import { Badge, Empty, Loading } from "../../components/ui";
+import { useSetPageLabel } from "../../hooks/useNavCollapseCtx";
 import { MessageList } from "../../components/chat/MessageList";
 import { CodeProjectPicker } from "../../components/code/CodeProjectPicker";
 import { CodeSessionList } from "../../components/code/CodeSessionList";
@@ -209,6 +209,11 @@ export function CodeScreen() {
 
   const hasProjects = !projects.isLoading && projectList.length > 0;
   const turns: CodeTurn[] = useMemo(() => msgs as unknown as CodeTurn[], [msgs]);
+  const activeTitle = useMemo(
+    () => sessions.data?.find((s) => s.id === sid)?.title || "",
+    [sessions.data, sid],
+  );
+  useSetPageLabel(activeTitle);
 
   // Detect unanswered ask_questions in the last assistant turn. Local "dismissed"
   // ref keys off the turn id so the panel re-appears for a fresh batch.
@@ -221,25 +226,17 @@ export function CodeScreen() {
   };
 
   return (
-    <div className="flex h-full min-h-0 flex-col overflow-hidden p-4" data-testid="screen-code">
-      <header className="mb-3 flex items-center justify-between gap-3">
-        <div className="min-w-0">
-          <h1 className="flex items-center gap-2 text-2xl font-bold tracking-tight">
-            <Code2 size={22} /> {t("code_module.title")}
-          </h1>
-          <p className="text-sm text-muted-fg">{t("code_module.desc")}</p>
-        </div>
-        <Badge tone="success">{t("code_module.badge")}</Badge>
-      </header>
-
+    <div className="flex h-full min-h-0" data-testid="screen-code">
       {projects.isLoading ? (
         <Loading />
       ) : !hasProjects ? (
-        <Empty>{t("code_module.no_projects")}</Empty>
+        <div className="grid flex-1 place-items-center">
+          <Empty>{t("code_module.no_projects")}</Empty>
+        </div>
       ) : (
-        <div className="flex min-h-0 flex-1 overflow-hidden rounded-xl border border-border bg-card/40">
-          {/* Left rail: project picker + session list */}
-          <aside className="flex w-60 shrink-0 flex-col border-r border-border">
+        <>
+          {/* Left: project picker + session list */}
+          <aside className="flex w-52 shrink-0 flex-col border-r border-border">
             <div className="shrink-0 border-b border-border p-2">
               <CodeProjectPicker
                 projects={projectList}
@@ -257,6 +254,9 @@ export function CodeScreen() {
               onRename={onRenameSession}
               onDelete={onDeleteSession}
             />
+            <div className="shrink-0 border-t border-border px-3 py-2">
+              <Badge tone="success">{t("code_module.badge")}</Badge>
+            </div>
           </aside>
 
           {/* Center: transcript + composer */}
@@ -300,7 +300,7 @@ export function CodeScreen() {
           </main>
 
           {/* Right: context + changes */}
-          <aside className="hidden w-80 shrink-0 flex-col border-l border-border lg:flex">
+          <aside className="hidden w-72 shrink-0 flex-col border-l border-border lg:flex">
             <CodeSidePanel
               pid={pid}
               turns={turns}
@@ -309,7 +309,7 @@ export function CodeScreen() {
               onRefreshChanges={() => void changes.mutate()}
             />
           </aside>
-        </div>
+        </>
       )}
     </div>
   );

package/src/interfaces/web/src/screens/modules/DeckScreen.tsx

@@ -76,24 +76,6 @@ export function DeckScreen() {
 
   return (
     <div className="mx-auto max-w-4xl space-y-6 p-6" data-testid="screen-deck">
-      <header className="flex items-start justify-between gap-4">
-        <div>
-          <h1 className="text-2xl font-bold tracking-tight">Deck</h1>
-          <p className="text-sm text-muted-fg">
-            App companion · widgets y escritorios.
-          </p>
-        </div>
-        <Button
-          size="sm"
-          variant="ghost"
-          onClick={() => mutate()}
-          disabled={isLoading}
-          title="Recargar manifest"
-        >
-          <RefreshCw size={14} className={isLoading ? "animate-spin" : ""} />
-        </Button>
-      </header>
-
       {/* Daemon info card */}
       {data && <DaemonCard manifest={data} />}
 
@@ -107,6 +89,11 @@ export function DeckScreen() {
             ? "Error al cargar el manifest."
             : `${widgets.length} widgets · ${enabledCount} externos habilitados`
         }
+        action={
+          <Button size="sm" variant="ghost" onClick={() => mutate()} disabled={isLoading} title="Recargar manifest">
+            <RefreshCw size={14} className={isLoading ? "animate-spin" : ""} />
+          </Button>
+        }
       >
         {isLoading && <Loading label="Cargando manifest del Deck…" />}
 

package/src/interfaces/web/src/screens/modules/DesktopScreen.tsx

@@ -96,15 +96,8 @@ export function DesktopScreen() {
 
   return (
     <div className="mx-auto max-w-6xl space-y-6 p-6" data-testid="screen-desktop">
-      <header>
-        <h1 className="text-2xl font-bold tracking-tight">Escritorio</h1>
-        <p className="text-sm text-muted-fg">
-          Ventana flotante de voz (Electron): atajo global, escucha por micrófono y muestra el chat.
-        </p>
-      </header>
-
       {/* ── Two-column layout: config on the left, last conversation on the right. ── */}
-      <div className="grid gap-6 lg:grid-cols-[1fr_1fr]">
+      <div className="grid gap-6 xl:grid-cols-[1fr_1fr]">
         {/* ── LEFT: configuration + status ─────────────────────────────── */}
         <div className="space-y-6">
           <Section title="Estado" description="La ventana se lanza desde la terminal o por autostart.">

package/src/interfaces/web/src/screens/modules/VoiceScreen.tsx

@@ -119,48 +119,47 @@ export function VoiceScreen() {
   };
 
   return (
-    <div className="mx-auto max-w-4xl space-y-6 p-6" data-testid="screen-voice">
-      <header>
-        <h1 className="text-2xl font-bold tracking-tight">Voces</h1>
-        <p className="text-sm text-muted-fg">
-          Configurá el texto a voz (TTS) y la transcripción (STT), elegí el motor por defecto y probá la voz.
-        </p>
-      </header>
+    <div className="mx-auto max-w-6xl p-6" data-testid="screen-voice">
+      <div className="grid gap-6 xl:grid-cols-2">
+        {/* Left: TTS providers */}
+        <Section
+          title="Proveedores de voz (TTS)"
+          description="Motores de síntesis. El estado lo reporta el daemon en vivo. Elegí cuál usar por defecto."
+        >
+          {provLoading || cfgLoading ? (
+            <Loading />
+          ) : provError ? (
+            <Empty>No se pudieron cargar los proveedores: {(provError as Error).message}</Empty>
+          ) : (
+            <VoiceProviderList
+              engines={engines}
+              order={order}
+              mode={mode}
+              configuredProvider={configuredProvider}
+              onSetMode={setMode}
+              onSetDefault={setDefault}
+              onToggleEnabled={toggleEnabled}
+              onReorder={reorder}
+              onConfigure={(id) => setEditing(id)}
+              busy={busyDefault}
+            />
+          )}
+        </Section>
 
-      <Section
-        title="Proveedores de voz (TTS)"
-        description="Motores de síntesis. El estado lo reporta el daemon en vivo. Elegí cuál usar por defecto."
-      >
-        {provLoading || cfgLoading ? (
-          <Loading />
-        ) : provError ? (
-          <Empty>No se pudieron cargar los proveedores: {(provError as Error).message}</Empty>
-        ) : (
-          <VoiceProviderList
-            engines={engines}
-            order={order}
-            mode={mode}
-            configuredProvider={configuredProvider}
-            onSetMode={setMode}
-            onSetDefault={setDefault}
-            onToggleEnabled={toggleEnabled}
-            onReorder={reorder}
-            onConfigure={(id) => setEditing(id)}
-            busy={busyDefault}
-          />
-        )}
-      </Section>
+        {/* Right: test + STT */}
+        <div className="space-y-6">
+          <Section title="Probar voz" description='Elegí con qué motor sintetizar y, si aplica, cómo querés que hable.'>
+            <VoiceTestCard engines={engines} defaultProvider={configuredProvider} mode={mode} />
+          </Section>
 
-      <Section title="Probar voz" description='Elegí con qué motor sintetizar y, si aplica, cómo querés que hable.'>
-        <VoiceTestCard engines={engines} defaultProvider={configuredProvider} mode={mode} />
-      </Section>
-
-      <Section
-        title="Transcripción (STT)"
-        description="Motor de voz a texto que usan el deck, Telegram y la CLI al escuchar."
-      >
-        {cfgLoading ? <Loading /> : <VoiceSttCard config={transcriptionCfg} onPatch={patchStt} />}
-      </Section>
+          <Section
+            title="Transcripción (STT)"
+            description="Motor de voz a texto que usan el deck, Telegram y la CLI al escuchar."
+          >
+            {cfgLoading ? <Loading /> : <VoiceSttCard config={transcriptionCfg} onPatch={patchStt} />}
+          </Section>
+        </div>
+      </div>
 
       <VoiceProviderModal
         open={!!editing}

package/src/interfaces/web/src/screens/project/ChatTab.tsx

@@ -12,6 +12,7 @@ import { InlineAskPanel, pendingAskQuestions } from "../../components/chat/Inlin
 import { useChat } from "../../hooks/useChat";
 import { useToast } from "../../components/Toast";
 import { t } from "../../i18n";
+import { usePersonaName } from "../../hooks/usePersonaName";
 import type { AgentEntry } from "../../types/daemon";
 
 // Virtual entry slug used in the agent dropdown to address the daemon-level
@@ -28,18 +29,19 @@ export function ChatTab({ pid }: { pid: string }) {
   const [model, setModel] = useState("");
   const [dismissedAskKey, setDismissedAskKey] = useState<string | null>(null);
   const { msgs, send: sendChat, stop, clear, streaming } = useChat(pid, (m) => toast.error(m));
+  const persona = usePersonaName();
 
   const agentList = agents.data || [];
-  // Virtual options shown in the dropdown — Roby is always first, then the
-  // real project agents. Roby works on every project (calls /projects/:pid
-  // /super-agent/chat) so we expose it everywhere, not just /base.
+  // Virtual options shown in the dropdown — the super-agent is always first,
+  // then the real project agents. It works on every project (calls
+  // /projects/:pid/super-agent/chat) so we expose it everywhere, not just /base.
   const isRoby = (slug: string | null | undefined) => slug === ROBY_SLUG;
   const dropdownOptions = useMemo(
     () => [
-      { value: ROBY_SLUG, label: "Roby (super-agent)" },
+      { value: ROBY_SLUG, label: `${persona} (super-agent)` },
       ...agentList.map((a) => ({ value: a.slug, label: a.slug })),
     ],
-    [agentList],
+    [agentList, persona],
   );
   const activeAgent = useMemo(
     () => agentList.find((a) => a.slug === activeSlug) || agentList[0],
@@ -75,10 +77,11 @@ export function ChatTab({ pid }: { pid: string }) {
 
   if (agents.isLoading) return <Loading />;
 
-  // Header subtitle differs: with Roby selected the chat goes through the
-  // super-agent (it CAN call tools); a project agent is a direct LLM call.
+  // Header subtitle differs: with the super-agent selected the chat goes
+  // through the super-agent loop (it CAN call tools); a project agent is a
+  // direct LLM call.
   const headerSubtitle = activeIsRoby
-    ? t("project.chat.roby_subtitle")
+    ? t("project.chat.superagent_subtitle", { persona })
     : t("project.chat.subtitle");
 
   return (
@@ -86,7 +89,7 @@ export function ChatTab({ pid }: { pid: string }) {
       <header className="flex shrink-0 items-center justify-between gap-3 border-b border-border px-4 py-3">
         <div className="min-w-0">
           <h2 className="text-sm font-semibold">
-            {activeIsRoby ? t("project.chat.roby_title") : t("project.chat.title")}
+            {activeIsRoby ? t("project.chat.superagent_title", { persona }) : t("project.chat.title")}
           </h2>
           <p className="truncate text-[11px] text-muted-fg">{headerSubtitle}</p>
         </div>

package/src/skills/apc-context/SKILL.md

@@ -0,0 +1,159 @@
+---
+name: apc-context
+description: "ALWAYS activate when the project has a .apc/ directory or AGENTS.md file. Do not wait to be asked. Read .apc/ before making any assumption about agents, memory, or project structure. Activate on: .apc/, AGENTS.md, 'which agents', 'list agents', 'agent context', 'who are the agents', any question about agents or memory in this project. IMPORTANT: if .apc/migrate.md exists, open the conversation with a migration offer before answering anything else. If the user declines, delete .apc/migrate.md immediately so it is not shown again."
+homepage: https://github.com/agentprojectcontext/agentprojectcontext
+---
+
+# Agent Project Context
+
+This project uses APC. APC stores portable project context in `.apc/` and `AGENTS.md`.
+
+APC does not store raw runtime sessions. Sessions, conversations, messages, caches, provider
+threads, and private runtime memory stay in the IDE, CLI, daemon, or user-level store that created
+them.
+
+## FIRST: check for pending migration
+
+Before doing anything else, check if `.apc/migrate.md` exists:
+
+```bash
+cat .apc/migrate.md 2>/dev/null
+```
+
+If it exists, open with this offer before answering anything else:
+
+> I see this project was initialized with Agent Project Context (APC).
+>
+> I found context files that may need migration:
+> [list files from .apc/migrate.md]
+>
+> I can read them, separate durable project context from runtime/private state, and migrate only
+> what belongs in APC.
+>
+> Want me to start?
+
+If the user says no or later, delete `.apc/migrate.md` so the offer is not repeated.
+
+## Migration rule: think, do not copy
+
+Read detected files first. Also read `AGENTS.md` if it exists.
+
+Classify content:
+
+| Content | Action |
+|---|---|
+| Agent definitions: name, model, description | Put in `.apc/agents/<name>.md` and/or `AGENTS.md` |
+| Shared project rules, stack notes, commands, testing policy | Keep in `AGENTS.md` |
+| Reusable instruction blocks | Move to `.apc/skills/<name>.md` |
+| Durable safe facts useful to all contributors | Add to `.apc/agents/<name>/memory.md` only after curation |
+| MCP expectations without secrets | Add to `.apc/mcps.json` |
+| Raw sessions, transcripts, conversations, messages, tool logs | Do not move into `.apc/`; leave with source runtime |
+| Secrets, tokens, credentials, private headers | Do not store in repository |
+| IDE UI settings or personal aliases | Leave in IDE/user config |
+| Instructions to store sessions under `.apc/` | Drop as obsolete |
+
+After migration:
+
+1. Update `AGENTS.md` as the root project contract.
+2. Create or update `.apc/agents/`, `.apc/skills/`, `.apc/mcps.json` as needed.
+3. Do not create `.apc/**/sessions`, `.apc/messages`, or `.apc/conversations`.
+4. Delete obsolete source files only when their useful project context was migrated or intentionally dropped.
+5. Delete `.apc/migrate.md`.
+6. Summarize what moved, what stayed local, and what was dropped.
+
+## APC structure
+
+```text
+AGENTS.md                        ← root project contract
+.apc/
+  project.json                   ← project metadata
+  .gitignore                     ← safety guard
+  agents/<name>.md               ← agent definition
+  agents/<name>/memory.md        ← optional curated project memory
+  skills/<name>.md               ← reusable project instructions
+  mcps.json                      ← MCP hints without secrets
+```
+
+Do not store:
+
+```text
+.apc/agents/<name>/sessions/
+.apc/sessions/
+.apc/conversations/
+.apc/messages/
+.apc/cache/
+.apc/tmp/
+.apc/private/
+.apc/secrets/
+```
+
+## Visibility rules
+
+| Data | Visibility | Commit? |
+|---|---|---|
+| Agent definitions, skills, project rules | `stable` / `project` | Yes |
+| Curated safe `memory.md` | `project` | Yes, if team-safe |
+| MCP hints without secrets | `project` | Yes |
+| Sessions, conversations, messages | `local` | No; runtime-owned |
+| Secrets, tokens, `*.secret.json`, `*.env` | `private` | Never |
+| Caches, temp files, databases | `ephemeral` | No |
+
+## Operating rules
+
+1. Read `AGENTS.md` and relevant `.apc/` files before assuming project context.
+2. Read agent definitions from `.apc/agents/<name>.md` when present.
+3. Read curated project memory from `.apc/agents/<name>/memory.md` when present.
+4. Write only durable, safe, curated facts to APC memory.
+5. Never write raw sessions, transcripts, messages, conversations, or tool logs into `.apc/`.
+6. Keep secrets out of APC and out of git.
+7. Treat `.apc/mcps.json` as MCP configuration hints, not as an MCP implementation.
+
+## Normalization
+
+If agent formats are broken or use legacy fields (role, skills in YAML), offer to normalize:
+
+```yaml
+---
+name: agent-name
+model: inherit
+description: Semantic activation trigger
+---
+```
+
+Identify and fix inconsistencies in `model` (use technical IDs or `inherit`) and ensure `description` is present for semantic activation.
+
+## Sessions
+
+Sessions belong to the runtime that created them.
+
+Examples:
+
+```text
+Codex runtime storage
+Claude Code runtime storage
+OpenCode runtime storage
+~/.apx/projects/<project-id>/agents/<name>/sessions/
+```
+
+At task end, provide the user a concise result. If project memory should be updated, write a short
+sanitized fact to `.apc/agents/<name>/memory.md` only when useful and safe.
+
+## APX
+
+Read `.apc/project.json` if present. It may contain an `apx` field:
+
+- `"installed"`: APX is available; use `apx` commands when useful.
+- `"declined"`: user chose not to install; do not suggest or run APX.
+- missing or `null`: unknown; do not assume APX is available.
+
+If APX is installed, it may manage runtime state outside the repository:
+
+```text
+~/.apx/projects/<project-id>/
+```
+
+APX can provide a local daemon, MCP management, Telegram bridge, routines, and runtime dispatch
+across Codex, Claude Code, OpenCode, Aider, or direct LLM engines. Those are APX runtime features,
+not APC portable-core requirements.
+
+Never use APX to write secrets or raw sessions into `.apc/`.

package/src/core/agent/ghost-guard.js

@@ -1,24 +0,0 @@
-export function isShortConfirmation(text) {
-  return /^(yes|y|si|si dale|dale|ok|okay|confirm|confirmed|go|proceed|do it)\b/i
-    .test(String(text || "").trim());
-}
-
-export function lastAssistantAskedForConfirmation(messages) {
-  for (let i = messages.length - 1; i >= 0; i--) {
-    if (messages[i]?.role !== "assistant") continue;
-    return /\b(confirm|confirmation|ok|okay|permission|allowed|proceed|do it|dale)\b/i.test(messages[i].content || "");
-  }
-  return false;
-}
-
-export function isGhostResponse(text) {
-  const t = String(text || "").trim();
-  if (t.length > 200) return false;
-  return /^(ok|okay|got it|understood|sure|of course|on it|dale|entendido|claro|voy|ya lo hago|dame un (segundo|momento)|un momento|let me|i (will|can|shall)|i'm (going|about)|give me a|ahora lo|enseguida|checking|looking|fetching|working on|stand by|please wait|un seg|dame sec)[\s.,!]*/i
-    .test(t);
-}
-
-export function looksLikeActionRequest(text) {
-  const t = String(text || "").trim().toLowerCase();
-  return /\b(list|show|find|get|fetch|search|run|execute|create|add|make|start|stop|delete|update|send|check|read|write|look|tell me|dame|mostra|busca|ejecuta|crea|agrega|mandá|revisá|corré|borrá|arrancá)\b/.test(t);
-}

package/src/core/agent/prompts/channels/terminal.md

@@ -1,16 +0,0 @@
-# Channel context
-Channel: **terminal** (interactive APX sys session) — the same OpenCode-style coding surface as the web Code module, just in the terminal. You can read, search, edit files and run shell commands.
-
-- CWD: {{cwd}}
-- References to "this directory", "this project", "here", "current folder" mean the CWD above — use it as the path argument; do not ask for a path
-
-Working style — KEEP GOING UNTIL THE TASK IS DONE (build mode):
-- You are an autonomous coding agent. Once the user gives a task, complete the WHOLE thing in this turn: chain as many tool calls as needed (read → edit → run → verify), do not stop after one or two steps.
-- NEVER stop to ask "do you want me to…?" / "¿confirmás…?" / "should I continue?". You already have permission on this surface — just do it. Only ask if the task is truly ambiguous and you genuinely cannot proceed.
-- NEVER announce an action and then end your turn ("now I'll edit the file." → stop). If you say you will do something, immediately call the tool and actually do it in the same turn.
-- After each tool result, decide the next concrete step and take it. Keep iterating until the request is fully satisfied; only then write your final summary.
-- If something fails, read the error, fix it, and retry — don't hand the problem back to the user.
-
-Formatting:
-- Markdown OK; keep readable; use code diffs when editing.
-- Lead with the result; keep prose tight. Don't re-paste full tool output the user can already see.