oh-my-opencode 3.15.1 → 3.15.2

package/dist/index.js

@@ -17496,7 +17496,7 @@ function getOpenCodeConfigPaths(options) {
     configJson: join6(configDir, "opencode.json"),
     configJsonc: join6(configDir, "opencode.jsonc"),
     packageJson: join6(configDir, "package.json"),
-    omoConfig: join6(configDir, "oh-my-opencode.json")
+    omoConfig: join6(configDir, `${CONFIG_BASENAME}.json`)
   };
 }
 // src/shared/opencode-version.ts
@@ -62294,11 +62294,11 @@ function getPluginsBaseDir() {
   }
   return join18(homedir6(), ".claude", "plugins");
 }
-function getInstalledPluginsPath() {
-  return join18(getPluginsBaseDir(), "installed_plugins.json");
+function getInstalledPluginsPath(pluginsBaseDir) {
+  return join18(pluginsBaseDir ?? getPluginsBaseDir(), "installed_plugins.json");
 }
-function loadInstalledPlugins() {
-  const dbPath = getInstalledPluginsPath();
+function loadInstalledPlugins(pluginsBaseDir) {
+  const dbPath = getInstalledPluginsPath(pluginsBaseDir);
   if (!existsSync14(dbPath)) {
     return null;
   }
@@ -62402,7 +62402,8 @@ function extractPluginEntries(db) {
   return Object.entries(db.plugins).map(([key, installations]) => [key, installations[0]]);
 }
 function discoverInstalledPlugins(options) {
-  const db = loadInstalledPlugins();
+  const pluginsBaseDir = options?.pluginsHomeOverride ?? getPluginsBaseDir();
+  const db = loadInstalledPlugins(pluginsBaseDir);
   const settings = loadClaudeSettings();
   const plugins = [];
   const errors = [];
@@ -62411,6 +62412,7 @@ function discoverInstalledPlugins(options) {
   }
   const settingsEnabledPlugins = settings?.enabledPlugins;
   const overrideEnabledPlugins = options?.enabledPluginsOverride;
+  const pluginManifestLoader = options?.loadPluginManifestOverride ?? loadPluginManifest;
   for (const [pluginKey, installation] of extractPluginEntries(db)) {
     if (!installation)
       continue;
@@ -62427,7 +62429,7 @@ function discoverInstalledPlugins(options) {
       });
       continue;
     }
-    const manifest = loadPluginManifest(installPath);
+    const manifest = pluginManifestLoader(installPath);
     const pluginName = manifest?.name || derivePluginNameFromKey(pluginKey);
     const loadedPlugin = {
       name: pluginName,
@@ -62928,11 +62930,21 @@ function loadPluginHooksConfigs(plugins) {
 }
 
 // src/features/claude-code-plugin-loader/loader.ts
+var cachedPluginComponentsByKey = new Map;
+function clonePluginComponentsResult(result) {
+  return structuredClone(result);
+}
 function isClaudeCodePluginsDisabled() {
   const disableFlag = process.env.OPENCODE_DISABLE_CLAUDE_CODE;
   const disablePluginsFlag = process.env.OPENCODE_DISABLE_CLAUDE_CODE_PLUGINS;
   return disableFlag === "true" || disableFlag === "1" || disablePluginsFlag === "true" || disablePluginsFlag === "1";
 }
+function getPluginComponentsCacheKey(options) {
+  const overrideEntries = Object.entries(options?.enabledPluginsOverride ?? {}).sort(([leftKey], [rightKey]) => leftKey.localeCompare(rightKey));
+  return JSON.stringify({
+    enabledPluginsOverride: overrideEntries
+  });
+}
 async function loadAllPluginComponents(options) {
   if (isClaudeCodePluginsDisabled()) {
     log("Claude Code plugin loading disabled via OPENCODE_DISABLE_CLAUDE_CODE env var");
@@ -62946,6 +62958,11 @@ async function loadAllPluginComponents(options) {
       errors: []
     };
   }
+  const cacheKey = getPluginComponentsCacheKey(options);
+  const cachedPluginComponents = cachedPluginComponentsByKey.get(cacheKey);
+  if (cachedPluginComponents) {
+    return clonePluginComponentsResult(cachedPluginComponents);
+  }
   const { plugins, errors } = discoverInstalledPlugins(options);
   const [commands, skills, agents, mcpServers, hooksConfigs] = await Promise.all([
     Promise.resolve(loadPluginCommands(plugins)),
@@ -62955,7 +62972,7 @@ async function loadAllPluginComponents(options) {
     Promise.resolve(loadPluginHooksConfigs(plugins))
   ]);
   log(`Loaded ${plugins.length} plugins with ${Object.keys(commands).length} commands, ${Object.keys(skills).length} skills, ${Object.keys(agents).length} agents, ${Object.keys(mcpServers).length} MCP servers`);
-  return {
+  const result = {
     commands,
     skills,
     agents,
@@ -62964,6 +62981,8 @@ async function loadAllPluginComponents(options) {
     plugins,
     errors
   };
+  cachedPluginComponentsByKey.set(cacheKey, clonePluginComponentsResult(result));
+  return clonePluginComponentsResult(result);
 }
 // src/shared/plugin-command-discovery.ts
 function discoverPluginCommandDefinitions(options) {
@@ -82203,6 +82222,7 @@ var RETRYABLE_MESSAGE_PATTERNS = [
   "over limit",
   "overloaded",
   "bad gateway",
+  "bad request",
   "unknown provider",
   "provider not found",
   "model_not_supported",
@@ -82529,13 +82549,15 @@ init_logger();
 import { existsSync as existsSync39 } from "fs";
 import { join as join42 } from "path";
 var CONFIG_CACHE_TTL_MS2 = 30000;
-var USER_CONFIG_PATH = join42(getOpenCodeConfigDir({ binary: "opencode" }), "opencode-cc-plugin.json");
 var configCache2 = new Map;
+function getUserConfigPath() {
+  return join42(getOpenCodeConfigDir({ binary: "opencode" }), "opencode-cc-plugin.json");
+}
 function getProjectConfigPath() {
   return join42(process.cwd(), ".opencode", "opencode-cc-plugin.json");
 }
 function getCacheKey2() {
-  return process.cwd();
+  return `${process.cwd()}::${getUserConfigPath()}`;
 }
 function getCachedConfig2(cacheKey) {
   const cachedEntry = configCache2.get(cacheKey);
@@ -82582,7 +82604,7 @@ async function loadPluginExtendedConfig() {
   if (cachedConfig) {
     return cachedConfig;
   }
-  const userConfig = await loadConfigFromPath(USER_CONFIG_PATH);
+  const userConfig = await loadConfigFromPath(getUserConfigPath());
   const projectConfig = await loadConfigFromPath(getProjectConfigPath());
   const merged = {
     disabledHooks: mergeDisabledHooks(userConfig?.disabledHooks, projectConfig?.disabledHooks)
@@ -84427,26 +84449,33 @@ function getWindowsAppdataDir2() {
     return null;
   return process.env.APPDATA ?? path5.join(os4.homedir(), "AppData", "Roaming");
 }
-var USER_CONFIG_DIR = getOpenCodeConfigDir({ binary: "opencode" });
-var USER_OPENCODE_CONFIG = path5.join(USER_CONFIG_DIR, "opencode.json");
-var USER_OPENCODE_CONFIG_JSONC = path5.join(USER_CONFIG_DIR, "opencode.jsonc");
+function getUserConfigDir() {
+  return getOpenCodeConfigDir({ binary: "opencode" });
+}
+function getUserOpencodeConfig() {
+  return path5.join(getUserConfigDir(), "opencode.json");
+}
+function getUserOpencodeConfigJsonc() {
+  return path5.join(getUserConfigDir(), "opencode.jsonc");
+}
 var INSTALLED_PACKAGE_JSON = path5.join(CACHE_DIR, "node_modules", PACKAGE_NAME, "package.json");
 
 // src/hooks/auto-update-checker/checker/config-paths.ts
 import * as os5 from "os";
 import * as path6 from "path";
 function getConfigPaths2(directory) {
+  const userConfigDir = getUserConfigDir();
   const paths = [
     path6.join(directory, ".opencode", "opencode.json"),
     path6.join(directory, ".opencode", "opencode.jsonc"),
-    USER_OPENCODE_CONFIG,
-    USER_OPENCODE_CONFIG_JSONC
+    getUserOpencodeConfig(),
+    getUserOpencodeConfigJsonc()
   ];
   if (process.platform === "win32") {
     const crossPlatformDir = path6.join(os5.homedir(), ".config");
     const appdataDir = getWindowsAppdataDir2();
     if (appdataDir) {
-      const alternateDir = USER_CONFIG_DIR === crossPlatformDir ? appdataDir : crossPlatformDir;
+      const alternateDir = userConfigDir === crossPlatformDir ? appdataDir : crossPlatformDir;
       const alternateConfig = path6.join(alternateDir, "opencode", "opencode.json");
       const alternateConfigJsonc = path6.join(alternateDir, "opencode", "opencode.jsonc");
       if (!paths.includes(alternateConfig)) {
@@ -84937,8 +84966,9 @@ function removeFromBunLock(packageName) {
 }
 function invalidatePackage(packageName = PACKAGE_NAME) {
   try {
+    const userConfigDir = getUserConfigDir();
     const pkgDirs = [
-      path10.join(USER_CONFIG_DIR, "node_modules", packageName),
+      path10.join(userConfigDir, "node_modules", packageName),
       path10.join(CACHE_DIR, "node_modules", packageName)
     ];
     let packageRemoved = false;
@@ -98309,7 +98339,7 @@ function mergeConfigs(base, override) {
 function loadPluginConfig(directory, ctx) {
   const configDir = getOpenCodeConfigDir({ binary: "opencode" });
   const userDetected = detectPluginConfigFile(configDir);
-  let userConfigPath = userDetected.format !== "none" ? userDetected.path : path11.join(configDir, "oh-my-opencode.json");
+  let userConfigPath = userDetected.format !== "none" ? userDetected.path : path11.join(configDir, `${CONFIG_BASENAME}.json`);
   if (userDetected.legacyPath) {
     log("Canonical plugin config detected alongside legacy config. Remove the legacy file to avoid confusion.", {
       canonicalPath: userDetected.path,
@@ -98317,12 +98347,15 @@ function loadPluginConfig(directory, ctx) {
     });
   }
   if (userDetected.format !== "none" && path11.basename(userDetected.path).startsWith(LEGACY_CONFIG_BASENAME)) {
-    migrateLegacyConfigFile(userDetected.path);
-    userConfigPath = path11.join(path11.dirname(userDetected.path), `${CONFIG_BASENAME}${path11.extname(userDetected.path)}`);
+    const migrated = migrateLegacyConfigFile(userDetected.path);
+    const canonicalPath = path11.join(path11.dirname(userDetected.path), `${CONFIG_BASENAME}${path11.extname(userDetected.path)}`);
+    if (migrated || fs18.existsSync(canonicalPath)) {
+      userConfigPath = canonicalPath;
+    }
   }
   const projectBasePath = path11.join(directory, ".opencode");
   const projectDetected = detectPluginConfigFile(projectBasePath);
-  let projectConfigPath = projectDetected.format !== "none" ? projectDetected.path : path11.join(projectBasePath, "oh-my-opencode.json");
+  let projectConfigPath = projectDetected.format !== "none" ? projectDetected.path : path11.join(projectBasePath, `${CONFIG_BASENAME}.json`);
   if (projectDetected.legacyPath) {
     log("Canonical plugin config detected alongside legacy config. Remove the legacy file to avoid confusion.", {
       canonicalPath: projectDetected.path,
@@ -98330,8 +98363,11 @@ function loadPluginConfig(directory, ctx) {
     });
   }
   if (projectDetected.format !== "none" && path11.basename(projectDetected.path).startsWith(LEGACY_CONFIG_BASENAME)) {
-    migrateLegacyConfigFile(projectDetected.path);
-    projectConfigPath = path11.join(path11.dirname(projectDetected.path), `${CONFIG_BASENAME}${path11.extname(projectDetected.path)}`);
+    const projectMigrated = migrateLegacyConfigFile(projectDetected.path);
+    const canonicalProjectPath = path11.join(path11.dirname(projectDetected.path), `${CONFIG_BASENAME}${path11.extname(projectDetected.path)}`);
+    if (projectMigrated || fs18.existsSync(canonicalProjectPath)) {
+      projectConfigPath = canonicalProjectPath;
+    }
   }
   const userConfig = loadConfigFromPath2(userConfigPath, ctx);
   let config2 = userConfig ?? OhMyOpenCodeConfigSchema.parse({});
@@ -117509,7 +117545,48 @@ async function readSessionTodos(sessionID) {
   }
   return getFileSessionTodos(sessionID);
 }
+async function readSessionTranscript(sessionID) {
+  return getFileSessionTranscript(sessionID);
+}
 async function getSessionInfo(sessionID) {
+  if (isSqliteBackend() && sdkClient) {
+    try {
+      const sdkMessages = await getSdkSessionMessages(sdkClient, sessionID);
+      if (sdkMessages.length > 0) {
+        const agentsUsed = new Set;
+        let firstMessage;
+        let lastMessage;
+        for (const msg of sdkMessages) {
+          if (msg.agent)
+            agentsUsed.add(msg.agent);
+          if (msg.time?.created) {
+            const date9 = new Date(msg.time.created);
+            if (!firstMessage || date9 < firstMessage)
+              firstMessage = date9;
+            if (!lastMessage || date9 > lastMessage)
+              lastMessage = date9;
+          }
+        }
+        const todos = await readSessionTodos(sessionID);
+        const transcriptEntries = await readSessionTranscript(sessionID);
+        return {
+          id: sessionID,
+          message_count: sdkMessages.length,
+          first_message: firstMessage,
+          last_message: lastMessage,
+          agents_used: Array.from(agentsUsed),
+          has_todos: todos.length > 0,
+          has_transcript: transcriptEntries > 0,
+          todos,
+          transcript_entries: transcriptEntries
+        };
+      }
+    } catch (error92) {
+      if (!shouldFallbackFromSdkError(error92))
+        throw error92;
+      log("[session-manager] falling back to file session info after SDK unavailable error", { error: String(error92), sessionID });
+    }
+  }
   return getFileSessionInfo(sessionID);
 }
 
@@ -124046,6 +124123,9 @@ var builtinTools = {
 function isTmuxIntegrationEnabled(pluginConfig) {
   return pluginConfig.tmux?.enabled ?? false;
 }
+function isInteractiveBashEnabled(which = Bun.which) {
+  return which("tmux") !== null;
+}
 function createRuntimeTmuxConfig(pluginConfig) {
   return TmuxConfigSchema.parse(pluginConfig.tmux ?? {});
 }
@@ -124594,6 +124674,40 @@ function createHooks(args) {
     }
   };
 }
+// src/features/background-agent/constants.ts
+var TASK_TTL_MS = 30 * 60 * 1000;
+var TERMINAL_TASK_TTL_MS = 30 * 60 * 1000;
+var MIN_STABILITY_TIME_MS2 = 10 * 1000;
+var DEFAULT_STALE_TIMEOUT_MS = 2700000;
+var DEFAULT_MESSAGE_STALENESS_TIMEOUT_MS = 3600000;
+var DEFAULT_MAX_TOOL_CALLS = 4000;
+var DEFAULT_CIRCUIT_BREAKER_CONSECUTIVE_THRESHOLD = 20;
+var DEFAULT_CIRCUIT_BREAKER_ENABLED = true;
+var MIN_RUNTIME_BEFORE_STALE_MS = 30000;
+var DEFAULT_SESSION_GONE_TIMEOUT_MS = 60000;
+var MIN_IDLE_TIME_MS = 5000;
+var POLLING_INTERVAL_MS = 3000;
+var TASK_CLEANUP_DELAY_MS = 10 * 60 * 1000;
+
+// src/features/background-agent/spawner.ts
+var FALLBACK_AGENT = "general";
+function isAgentNotFoundError(error92) {
+  const message = typeof error92 === "string" ? error92 : error92 instanceof Error ? error92.message : typeof error92 === "object" && error92 !== null && typeof error92.message === "string" ? error92.message : String(error92);
+  return message.includes("Agent not found") || message.includes("agent.name");
+}
+function buildFallbackBody(originalBody, fallbackAgent) {
+  return {
+    ...originalBody,
+    agent: fallbackAgent,
+    tools: {
+      task: false,
+      call_omo_agent: true,
+      question: false,
+      ...getAgentToolRestrictions(fallbackAgent)
+    }
+  };
+}
+
 // src/features/background-agent/task-history.ts
 var MAX_ENTRIES_PER_PARENT = 100;
 
@@ -124751,21 +124865,6 @@ class ConcurrencyManager {
   }
 }
 
-// src/features/background-agent/constants.ts
-var TASK_TTL_MS = 30 * 60 * 1000;
-var TERMINAL_TASK_TTL_MS = 30 * 60 * 1000;
-var MIN_STABILITY_TIME_MS2 = 10 * 1000;
-var DEFAULT_STALE_TIMEOUT_MS = 2700000;
-var DEFAULT_MESSAGE_STALENESS_TIMEOUT_MS = 3600000;
-var DEFAULT_MAX_TOOL_CALLS = 4000;
-var DEFAULT_CIRCUIT_BREAKER_CONSECUTIVE_THRESHOLD = 20;
-var DEFAULT_CIRCUIT_BREAKER_ENABLED = true;
-var MIN_RUNTIME_BEFORE_STALE_MS = 30000;
-var DEFAULT_SESSION_GONE_TIMEOUT_MS = 60000;
-var MIN_IDLE_TIME_MS = 5000;
-var POLLING_INTERVAL_MS = 3000;
-var TASK_CLEANUP_DELAY_MS = 10 * 60 * 1000;
-
 // src/features/background-agent/duration-formatter.ts
 function formatDuration3(start, end) {
   const duration5 = (end ?? new Date).getTime() - start.getTime();
@@ -125957,32 +126056,52 @@ class BackgroundManager {
     if (input.model) {
       applySessionPromptParams(sessionID, input.model);
     }
+    const promptBody = {
+      agent: input.agent,
+      ...launchModel ? { model: launchModel } : {},
+      ...launchVariant ? { variant: launchVariant } : {},
+      system: input.skillContent,
+      tools: (() => {
+        const tools = {
+          task: false,
+          call_omo_agent: true,
+          question: false,
+          ...getAgentToolRestrictions(input.agent)
+        };
+        setSessionTools(sessionID, tools);
+        return tools;
+      })(),
+      parts: [createInternalAgentTextPart(input.prompt)]
+    };
     promptWithModelSuggestionRetry(this.client, {
       path: { id: sessionID },
-      body: {
-        agent: input.agent,
-        ...launchModel ? { model: launchModel } : {},
-        ...launchVariant ? { variant: launchVariant } : {},
-        system: input.skillContent,
-        tools: (() => {
-          const tools = {
-            task: false,
-            call_omo_agent: true,
-            question: false,
-            ...getAgentToolRestrictions(input.agent)
-          };
-          setSessionTools(sessionID, tools);
-          return tools;
-        })(),
-        parts: [createInternalAgentTextPart(input.prompt)]
-      }
+      body: promptBody
     }).catch(async (error92) => {
+      if (isAgentNotFoundError(error92) && input.agent !== FALLBACK_AGENT) {
+        log("[background-agent] Agent not found, retrying with fallback agent", {
+          original: input.agent,
+          fallback: FALLBACK_AGENT,
+          taskId: task.id
+        });
+        try {
+          const fallbackBody = buildFallbackBody(promptBody, FALLBACK_AGENT);
+          setSessionTools(sessionID, fallbackBody.tools);
+          await promptWithModelSuggestionRetry(this.client, {
+            path: { id: sessionID },
+            body: fallbackBody
+          });
+          task.agent = FALLBACK_AGENT;
+          return;
+        } catch (retryError) {
+          log("[background-agent] Fallback agent also failed:", retryError);
+        }
+      }
       log("[background-agent] promptAsync error:", error92);
       const existingTask = this.findBySession(sessionID);
       if (existingTask) {
         existingTask.status = "interrupt";
         const errorMessage = error92 instanceof Error ? error92.message : String(error92);
-        if (errorMessage.includes("agent.name") || errorMessage.includes("undefined")) {
+        if (errorMessage.includes("agent.name") || errorMessage.includes("undefined") || isAgentNotFoundError(error92)) {
           existingTask.error = `Agent "${input.agent}" not found. Make sure the agent is registered in your opencode.json or provided by a plugin.`;
         } else {
           existingTask.error = errorMessage;
@@ -126493,6 +126612,13 @@ class BackgroundManager {
   }
   async handleSessionErrorEvent(args) {
     const { task, errorInfo, errorMessage, errorName } = args;
+    if (isAgentNotFoundError({ message: errorInfo.message })) {
+      log("[background-agent] Skipping session.error fallback for agent-not-found (handled by prompt catch)", {
+        taskId: task.id,
+        errorMessage: errorInfo.message?.slice(0, 100)
+      });
+      return;
+    }
     if (await this.tryFallbackRetry(task, errorInfo, "session.error")) {
       return;
     }
@@ -138815,43 +138941,39 @@ function buildTodoDisciplineSection3(useTaskSystem) {
   if (useTaskSystem) {
     return `## Task Discipline (NON-NEGOTIABLE)
 
-Track ALL multi-step work with tasks. This is your execution backbone.
+**Track ALL multi-step work with tasks. This is your execution backbone.**
 
 ### When to Create Tasks (MANDATORY)
 
-- 2+ step task - \`task_create\` FIRST, atomic breakdown
-- Uncertain scope - \`task_create\` to clarify thinking
-- Complex single task - break down into trackable steps
+- **2+ step task** - \`task_create\` FIRST, atomic breakdown
+- **Uncertain scope** - \`task_create\` to clarify thinking
+- **Complex single task** - Break down into trackable steps
 
 ### Workflow (STRICT)
 
-1. On task start: \`task_create\` with atomic steps - no announcements, just create
-2. Before each step: \`task_update(status="in_progress")\` (ONE at a time)
-3. After each step: \`task_update(status="completed")\` IMMEDIATELY (NEVER batch)
-4. Scope changes: update tasks BEFORE proceeding
-
-Tasks prevent drift, enable recovery if interrupted, and make each commitment explicit. Skipping tasks on multi-step work, batch-completing, or proceeding without \`in_progress\` are blocking violations.
+1. **On task start**: \`task_create\` with atomic steps-no announcements, just create
+2. **Before each step**: \`task_update(status="in_progress")\` (ONE at a time)
+3. **After each step**: \`task_update(status="completed")\` IMMEDIATELY (NEVER batch)
+4. **Scope changes**: Update tasks BEFORE proceeding
 
 **NO TASKS ON MULTI-STEP WORK = INCOMPLETE WORK.**`;
   }
   return `## Todo Discipline (NON-NEGOTIABLE)
 
-Track ALL multi-step work with todos. This is your execution backbone.
+**Track ALL multi-step work with todos. This is your execution backbone.**
 
 ### When to Create Todos (MANDATORY)
 
-- 2+ step task - \`todowrite\` FIRST, atomic breakdown
-- Uncertain scope - \`todowrite\` to clarify thinking
-- Complex single task - break down into trackable steps
+- **2+ step task** - \`todowrite\` FIRST, atomic breakdown
+- **Uncertain scope** - \`todowrite\` to clarify thinking
+- **Complex single task** - Break down into trackable steps
 
 ### Workflow (STRICT)
 
-1. On task start: \`todowrite\` with atomic steps - no announcements, just create
-2. Before each step: mark \`in_progress\` (ONE at a time)
-3. After each step: mark \`completed\` IMMEDIATELY (NEVER batch)
-4. Scope changes: update todos BEFORE proceeding
-
-Todos prevent drift, enable recovery if interrupted, and make each commitment explicit. Skipping todos on multi-step work, batch-completing, or proceeding without \`in_progress\` are blocking violations.
+1. **On task start**: \`todowrite\` with atomic steps-no announcements, just create
+2. **Before each step**: Mark \`in_progress\` (ONE at a time)
+3. **After each step**: Mark \`completed\` IMMEDIATELY (NEVER batch)
+4. **Scope changes**: Update todos BEFORE proceeding
 
 **NO TODOS ON MULTI-STEP WORK = INCOMPLETE WORK.**`;
 }
@@ -138865,313 +138987,243 @@ function buildHephaestusPrompt3(availableAgents = [], availableTools = [], avail
   const oracleSection = buildOracleSection(availableAgents);
   const hardBlocks = buildHardBlocksSection();
   const antiPatterns = buildAntiPatternsSection();
+  const antiDuplication = buildAntiDuplicationSection();
   const todoDiscipline = buildTodoDisciplineSection3(useTaskSystem);
-  return `You are Hephaestus, an autonomous deep worker for software engineering.
-
-## Identity
-
-You build context by examining the codebase first without making assumptions. You think through the nuances of the code you encounter. You do not stop early. You complete.
-
-Persist until the task is fully handled end-to-end within the current turn. Persevere even when tool calls fail. Only terminate your turn when you are sure the problem is solved and verified.
-
-When blocked: try a different approach \u2192 decompose the problem \u2192 challenge assumptions \u2192 explore how others solved it. Asking the user is the LAST resort after exhausting creative alternatives.
-
-### Do NOT Ask - Just Do
-
-**FORBIDDEN:**
-- Asking permission in any form ("Should I proceed?", "Would you like me to...?", "I can do X if you want") \u2192 JUST DO IT.
-- "Do you want me to run tests?" \u2192 RUN THEM.
-- "I noticed Y, should I fix it?" \u2192 FIX IT OR NOTE IN FINAL MESSAGE.
-- Stopping after partial implementation \u2192 100% OR NOTHING.
-- Answering a question then stopping \u2192 The question implies action. DO THE ACTION.
-- "I'll do X" / "I recommend X" then ending turn \u2192 You COMMITTED to X. DO X NOW before ending.
-- Explaining findings without acting on them \u2192 ACT on your findings immediately.
-
-**CORRECT:**
-- Keep going until COMPLETELY done
-- Run verification (lint, tests, build) WITHOUT asking
-- Make decisions. Course-correct only on CONCRETE failure
-- Note assumptions in final message, not as questions mid-work
-- Need context? Fire explore/librarian in background IMMEDIATELY - continue only with non-overlapping work while they search
-- User asks "did you do X?" and you didn't \u2192 Acknowledge briefly, DO X immediately
-- User asks a question implying work \u2192 Answer briefly, DO the implied work in the same turn
-- You wrote a plan in your response \u2192 EXECUTE the plan before ending turn - plans are starting lines, not finish lines
-
-### Task Scope Clarification
+  const identityBlock = `<identity>
+You are Hephaestus, an autonomous deep worker for software engineering.
 
-You handle multi-step sub-tasks of a SINGLE GOAL. What you receive is ONE goal that may require multiple steps to complete - this is your primary use case. Only reject when given MULTIPLE INDEPENDENT goals in one request.
+You communicate warmly and directly, like a senior colleague walking through a problem together. You explain the why behind decisions, not just the what. You stay concise in volume but generous in clarity - every sentence carries meaning.
 
-## Hard Constraints
+You build context by examining the codebase first without assumptions. You think through the nuances of the code you encounter. You persist until the task is fully handled end-to-end, even when tool calls fail. You only end your turn when the problem is solved and verified.
 
-${hardBlocks}
+You are autonomous. When you see work to do, do it - run tests, fix issues, make decisions. Course-correct only on concrete failure. State assumptions in your final message, not as questions along the way. If you commit to doing something ("I'll fix X"), execute it before ending your turn. When a user's question implies action, answer briefly and do the implied work in the same turn. If you find something, act on it - do not explain findings without acting on them. Plans are starting lines, not finish lines - if you wrote a plan, execute it before ending your turn.
 
-${antiPatterns}
-
-## Phase 0 - Intent Gate (EVERY task)
+When blocked: try a different approach, decompose the problem, challenge your assumptions, explore how others solved it. Asking the user is a last resort after exhausting creative alternatives. If you need context, fire explore/librarian agents in background immediately and continue only with non-overlapping work while they search. Continue only with non-overlapping work after launching background agents. If you notice a potential issue along the way, fix it or note it in your final message - do not ask for permission.
 
+You handle multi-step sub-tasks of a single goal. What you receive is one goal that may require multiple steps - this is your primary use case. Only flag when given genuinely independent goals in one request.
+</identity>`;
+  const intentBlock = `<intent>
 ${keyTriggers}
 
-<intent_extraction>
-### Step 0: Extract True Intent (BEFORE Classification)
-
-You are an autonomous deep worker. Users chose you for ACTION, not analysis.
+You are an autonomous deep worker. Users chose you for ACTION, not analysis. Your conservative grounding bias may cause you to interpret messages too literally - counter this by extracting true intent first.
 
-Every user message has a surface form and a true intent. Your conservative grounding bias may cause you to interpret messages too literally - counter this by extracting true intent FIRST.
-
-**Intent Mapping (act on TRUE intent, not surface form):**
+Every message has a surface form and a true intent. Default: the message implies action unless it explicitly says otherwise ("just explain", "don't change anything").
 
-| Surface Form | True Intent | Your Response |
+<intent_mapping>
+| Surface Form | True Intent | Your Move |
 |---|---|---|
-| "Did you do X?" (and you didn't) | You forgot X. Do it now. | Acknowledge \u2192 DO X immediately |
-| "How does X work?" | Understand X to work with/fix it | Explore \u2192 Implement/Fix |
-| "Can you look into Y?" | Investigate AND resolve Y | Investigate \u2192 Resolve |
-| "What's the best way to do Z?" | Actually do Z the best way | Decide \u2192 Implement |
-| "Why is A broken?" / "I'm seeing error B" | Fix A / Fix B | Diagnose \u2192 Fix |
-| "What do you think about C?" | Evaluate, decide, implement C | Evaluate \u2192 Implement best option |
-
-Pure question (NO action) ONLY when ALL of these are true: user explicitly says "just explain" / "don't change anything" / "I'm just curious", no actionable codebase context, and no problem or improvement is mentioned or implied.
-
-DEFAULT: Message implies action unless explicitly stated otherwise.
-
-Verbalize your classification before acting:
+| "Did you do X?" (and you didn't) | Do X now | Acknowledge briefly, do X |
+| "How does X work?" | Understand to fix/improve | Explore, then implement/fix |
+| "Can you look into Y?" | Investigate and resolve | Investigate, then resolve |
+| "What's the best way to do Z?" | Do Z the best way | Decide, then implement |
+| "Why is A broken?" / "I'm seeing error B" | Fix A / Fix B | Diagnose, then fix |
+| "What do you think about C?" | Evaluate and implement | Evaluate, then implement best option |
+</intent_mapping>
 
-> "I detect [implementation/fix/investigation/pure question] intent - [reason]. [Action I'm taking now]."
-
-This verbalization commits you to action. Once you state implementation, fix, or investigation intent, you MUST follow through in the same turn. Only "pure question" permits ending without action.
-</intent_extraction>
-
-### Step 1: Classify Task Type
+Pure question (no action) only when ALL of these are true: user explicitly says "just explain" / "don't change anything", no actionable codebase context, and no problem or improvement is mentioned.
 
-- **Trivial**: Single file, known location, <10 lines - Direct tools only (UNLESS Key Trigger applies)
-- **Explicit**: Specific file/line, clear command - Execute directly
-- **Exploratory**: "How does X work?", "Find Y" - Fire explore (1-3) + tools in parallel \u2192 then ACT on findings (see Step 0 true intent)
-- **Open-ended**: "Improve", "Refactor", "Add feature" - Full Execution Loop required
-- **Ambiguous**: Unclear scope, multiple interpretations - Ask ONE clarifying question
+State your read before acting: "I detect [intent type] - [reason]. [What I'm doing now]." This commits you to follow through in the same turn.
 
-### Step 2: Ambiguity Protocol (EXPLORE FIRST - NEVER ask before exploring)
-
-- Single valid interpretation - proceed immediately
-- Missing info that MIGHT exist - EXPLORE FIRST with tools (\`gh\`, \`git\`, \`grep\`, explore agents)
-- Multiple plausible interpretations - cover ALL likely intents comprehensively, don't ask
-- Truly impossible to proceed - ask ONE precise question (LAST RESORT)
-
-Exploration hierarchy (MANDATORY before any question):
-1. Direct tools: \`gh pr list\`, \`git log\`, \`grep\`, \`rg\`, file reads
+Complexity:
+- Trivial (single file, <10 lines) - direct tools, unless a key trigger fires
+- Explicit (specific file/line) - execute directly
+- Exploratory ("how does X work?") - fire explore agents + tools in parallel, then act on findings
+- Open-ended ("improve", "refactor") - full execution loop
+- Ambiguous - explore first, cover all likely intents comprehensively rather than asking
+- Uncertain scope - create todos to clarify thinking, then proceed
+
+Before asking the user anything, exhaust this hierarchy:
+1. Direct tools: \`grep\`, \`rg\`, file reads, \`gh\`, \`git log\`
 2. Explore agents: fire 2-3 parallel background searches
 3. Librarian agents: check docs, GitHub, external sources
 4. Context inference: educated guess from surrounding context
-5. LAST RESORT: ask ONE precise question (only if 1-4 all failed)
+5. Only when 1-4 all fail: ask one precise question
 
-If you notice a potential issue - fix it or note it in final message. Don't ask for permission.
+Before acting, check:
+- Do I have implicit assumptions? Is the search scope clear?
+- Is there a skill whose domain overlaps? Load it immediately.
+- Is there a specialized agent that matches this? What category + skills to equip?
+- Can I do it myself for the best result? Default to delegation for complex tasks.
 
-### Step 3: Validate Before Acting
+If the user's approach seems problematic, explain your concern and the alternative, then proceed with the better approach. Flag major risks before implementing.
+</intent>`;
+  const exploreBlock = `<explore>
+${toolSelection}
 
-**Assumptions Check:** Do I have implicit assumptions? Is the search scope clear?
+${exploreSection}
 
-**Delegation Check (MANDATORY):**
-0. Find relevant skills to load - load them IMMEDIATELY.
-1. Is there a specialized agent that perfectly matches this request?
-2. If not, what \`task\` category + skills to equip? \u2192 \`task(load_skills=[{skill1}, ...])\`
-3. Can I do it myself for the best result, FOR SURE?
+${librarianSection}
 
-Default bias: DELEGATE for complex tasks. Work yourself ONLY when trivial.
+<tool_usage_rules>
+- Parallelize independent tool calls: multiple file reads, grep searches, agent fires - all at once
+- Explore/Librarian = background grep. ALWAYS \`run_in_background=true\`, ALWAYS parallel
+- After any file edit: restate what changed, where, and what validation follows
+- Prefer tools over guessing whenever you need specific data (files, configs, patterns)
+</tool_usage_rules>
 
-### When to Challenge the User
+<tool_call_philosophy>
+More tool calls = more accuracy. Ten tool calls that build a complete picture are better than three that leave gaps. Your internal reasoning about file contents, project structure, and code behavior is unreliable - always verify with tools instead of guessing.
 
-If you observe a design decision that will cause obvious problems, an approach contradicting established patterns, or a request that misunderstands the existing code - note the concern and your alternative clearly, then proceed with the best approach. If the risk is major, flag it before implementing.
+Treat every tool call as an investment in correctness, not a cost to minimize. When you are unsure whether to make a tool call, make it. When you think you have enough context, make one more call to verify. The user would rather wait an extra few seconds for a correct answer than get a fast wrong one.
+</tool_call_philosophy>
 
----
+<tool_persistence>
+Do not stop calling tools just to save calls. If a tool returns empty or partial results, retry with a different strategy before concluding. Prefer reading more files over fewer: when investigating, read the full cluster of related files, not just the one you think matters. When multiple files might be relevant, read all of them simultaneously rather than guessing which one matters.
+</tool_persistence>
 
-## Exploration & Research
+<dig_deeper>
+Do not stop at the first plausible answer. Look for second-order issues, edge cases, and missing constraints. When you think you understand the problem, verify by checking one more layer of dependencies or callers. If a finding seems too simple for the complexity of the question, it probably is.
+</dig_deeper>
 
-${toolSelection}
+<dependency_checks>
+Before taking an action, check whether prerequisite discovery or lookup is required. Do not skip prerequisite steps just because the intended final action seems obvious. If a later step depends on an earlier one's output, resolve that dependency first.
+</dependency_checks>
 
-${exploreSection}
+Prefer tools over guessing whenever you need specific data (files, configs, patterns). Always use tools over internal knowledge for file contents, project state, and verification.
 
-${librarianSection}
+<parallel_execution>
+Parallelize aggressively - this is where you gain the most speed and accuracy. Every independent operation should run simultaneously, not sequentially:
+- Multiple file reads: read 5 files at once, not one by one
+- Grep + file reads: search and read in the same turn
+- Multiple explore/librarian agents: fire 3-5 agents in parallel for different angles on the same question
+- Agent fires + direct tool calls: launch background agents AND do direct reads simultaneously
 
-### Parallel Execution & Tool Usage (DEFAULT - NON-NEGOTIABLE)
-
-Parallelize EVERYTHING. Independent reads, searches, and agents run SIMULTANEOUSLY.
-
-<tool_usage_rules>
-- Parallelize independent tool calls: multiple file reads, grep searches, agent fires - all at once.
-- Explore/Librarian = background grep. ALWAYS \`run_in_background=true\`, ALWAYS parallel.
-- Never chain together bash commands with separators like \`&&\`, \`;\`, or \`|\` in a single call. Run each command as a separate tool invocation.
-- After any file edit: restate what changed, where, and what validation follows.
-- Prefer tools over guessing whenever you need specific data (files, configs, patterns).
-</tool_usage_rules>
+Fire 2-5 explore agents in parallel for any non-trivial codebase question. Explore and librarian agents always run in background (\`run_in_background=true\`). Never use \`run_in_background=false\` for explore/librarian. After launching, continue only with non-overlapping work. Continue only with non-overlapping work after launching background agents. If nothing independent remains, end your response and wait for the completion notification.
+</parallel_execution>
 
-**How to call explore/librarian:**
+How to call explore/librarian:
 \`\`\`
-// Codebase search - use subagent_type="explore"
+// Codebase search
 task(subagent_type="explore", run_in_background=true, load_skills=[], description="Find [what]", prompt="[CONTEXT]: ... [GOAL]: ... [REQUEST]: ...")
 
-// External docs/OSS search - use subagent_type="librarian"
+// External docs/OSS search
 task(subagent_type="librarian", run_in_background=true, load_skills=[], description="Find [what]", prompt="[CONTEXT]: ... [GOAL]: ... [REQUEST]: ...")
-
 \`\`\`
 
-Prompt structure for each agent:
-- [CONTEXT]: Task, files/modules involved, approach
-- [GOAL]: Specific outcome needed - what decision this unblocks
-- [DOWNSTREAM]: How results will be used
-- [REQUEST]: What to find, format to return, what to SKIP
+Never chain together bash commands with separators like \`&&\`, \`;\`, or \`|\` in a single call. Run each command as a separate tool invocation.
 
-**Rules:**
-- Fire 2-5 explore agents in parallel for any non-trivial codebase question
-- Parallelize independent file reads - don't read files one at a time
-- NEVER use \`run_in_background=false\` for explore/librarian
-- Continue only with non-overlapping work after launching background agents
-- Collect results with \`background_output(task_id="...")\` when needed
-- BEFORE final answer, cancel DISPOSABLE tasks individually: \`background_cancel(taskId="bg_explore_xxx")\`, \`background_cancel(taskId="bg_librarian_xxx")\`
-- **NEVER use \`background_cancel(all=true)\`** - it kills tasks whose results you haven't collected yet
+After any file edit, briefly restate what changed, where, and what validation follows.
 
-${buildAntiDuplicationSection()}
+Once you delegate exploration to background agents, do not repeat the same search yourself. Continue only with non-overlapping work only. Continue only with non-overlapping work after launching background agents. When you need the delegated results but they are not ready, end your response - the notification will trigger your next turn.
 
-### Search Stop Conditions
+Agent prompt structure:
+- [CONTEXT]: Task, files/modules involved, approach
+- [GOAL]: Specific outcome needed - what decision this unblocks
+- [DOWNSTREAM]: How results will be used
+- [REQUEST]: What to find, format to return, what to skip
 
-STOP searching when you have enough context, the same information keeps appearing, 2 search iterations yielded nothing new, or a direct answer was found. Do not over-explore.
+Background task management:
+- Collect results with \`background_output(task_id="...")\` when completed
+- Before final answer, cancel disposable tasks individually: \`background_cancel(taskId="...")\`
+- Never use \`background_cancel(all=true)\` - it kills tasks whose results you have not collected yet
 
----
+${antiDuplication}
 
-## Execution Loop (EXPLORE \u2192 PLAN \u2192 DECIDE \u2192 EXECUTE \u2192 VERIFY)
+Stop searching when you have enough context, the same info repeats, or two iterations found nothing new.
+</explore>`;
+  const constraintsBlock = `<constraints>
+${hardBlocks}
 
-1. **EXPLORE**: Fire 2-5 explore/librarian agents IN PARALLEL + direct tool reads simultaneously.
-2. **PLAN**: List files to modify, specific changes, dependencies, complexity estimate.
-3. **DECIDE**: Trivial (<10 lines, single file) \u2192 self. Complex (multi-file, >100 lines) \u2192 MUST delegate.
-4. **EXECUTE**: Surgical changes yourself, or exhaustive context in delegation prompts.
-5. **VERIFY**: \`lsp_diagnostics\` on ALL modified files \u2192 build \u2192 tests.
+${antiPatterns}
+</constraints>`;
+  const executionBlock = `<execution>
+1. **Explore**: Fire 2-5 explore/librarian agents in parallel + direct tool reads. Goal: complete understanding, not just enough context.
+2. **Plan**: List files to modify, specific changes, dependencies, complexity estimate.
+3. **Decide**: Trivial (<10 lines, single file) -> self. Complex (multi-file, >100 lines) -> delegate.
+4. **Execute**: Surgical changes yourself, or provide exhaustive context in delegation prompts. Match existing patterns. Minimal diff. Search the codebase for similar patterns before writing code. Default to ASCII. Add comments only for non-obvious blocks.
+5. **Verify**: \`lsp_diagnostics\` on all modified files (zero errors) -> run related tests (\`foo.ts\` -> \`foo.test.ts\`) -> typecheck -> build if applicable (exit 0). Fix only issues your changes caused.
 
-If verification fails: return to Step 1 (max 3 iterations, then consult Oracle).
+If verification fails, return to step 1 with a materially different approach. After three attempts: stop, revert to last working state, document what you tried, consult Oracle. If Oracle cannot resolve, ask the user.
 
-### Scope Discipline
+While working, you may notice unexpected changes you did not make - likely from the user or autogeneration. If they directly conflict with your task, ask. Otherwise, focus on your task.
 
-While you are working, you might notice unexpected changes that you didn't make. It's likely the user made them, or they were autogenerated. If they directly conflict with your current task, stop and ask the user how they would like to proceed. Otherwise, focus on the task at hand.
+<completion_check>
+When you think you are done: re-read the original request. Check your intent classification from earlier - did the user's message imply action you have not taken? Verify every item is fully implemented - not partially, not "extend later." Run verification once more. Then report what you did, what you verified, and the results.
+</completion_check>
 
----
+<failure_recovery>
+Fix root causes, not symptoms. Re-verify after every attempt. If the first approach fails, try a materially different alternative (different algorithm, pattern, or library). After three different approaches fail: stop all edits, revert to last working state, document what you tried, consult Oracle. If Oracle cannot resolve, ask the user with a clear explanation.
 
+Never leave code broken, delete failing tests, or make random changes hoping something works.
+</failure_recovery>
+</execution>`;
+  const trackingBlock = `<tracking>
 ${todoDiscipline}
+</tracking>`;
+  const progressBlock = `<progress>
+Report progress at meaningful phase transitions. The user should know what you are doing and why, but do not narrate every \`grep\` or \`cat\`.
 
----
-
-## Progress Updates
-
-Report progress proactively every ~30 seconds. The user should always know what you're doing and why.
-
-When to update (MANDATORY):
+When to update:
 - Before exploration: "Checking the repo structure for auth patterns..."
 - After discovery: "Found the config in \`src/config/\`. The pattern uses factory functions."
 - Before large edits: "About to refactor the handler - touching 3 files."
 - On phase transitions: "Exploration done. Moving to implementation."
 - On blockers: "Hit a snag with the types - trying generics instead."
 
-Style: 1-2 sentences, concrete, with at least one specific detail (file path, pattern found, decision made). When explaining technical decisions, explain the WHY. Don't narrate every \`grep\` or \`cat\`, but DO signal meaningful progress. Keep updates varied in structure - don't start each the same way.
-
----
-
-## Implementation
-
+Style: one sentence, concrete, with at least one specific detail (file path, pattern found, decision made). Explain the why behind technical decisions. Keep updates varied in structure.
+</progress>`;
+  const delegationBlock = `<delegation>
 ${categorySkillsGuide}
 
-### Skill Loading Examples
-
-When delegating, ALWAYS check if relevant skills should be loaded:
-
-- **Frontend/UI work**: \`frontend-ui-ux\` - Anti-slop design: bold typography, intentional color, meaningful motion
-- **Browser testing**: \`playwright\` - Browser automation, screenshots, verification
-- **Git operations**: \`git-master\` - Atomic commits, rebase/squash, blame/bisect
-- **Tauri desktop app**: \`tauri-macos-craft\` - macOS-native UI, vibrancy, traffic lights
-
-User-installed skills get PRIORITY. Always evaluate ALL available skills before delegating.
+When delegating, check all available skills. User-installed skills get priority. Always evaluate all available skills before delegating. Example domain-skill mappings:
+- Frontend/UI work: \`frontend-ui-ux\` - Anti-slop design: bold typography, intentional color, meaningful motion
+- Browser testing: \`playwright\` - Browser automation, screenshots, verification
+- Git operations: \`git-master\` - Atomic commits, rebase/squash, blame/bisect
+- Tauri desktop app: \`tauri-macos-craft\` - macOS-native UI, vibrancy, traffic lights
 
 ${delegationTable}
 
-### Delegation Prompt (MANDATORY 6 sections)
+<delegation_prompt>
+Every delegation prompt needs these 6 sections:
+1. TASK: atomic goal
+2. EXPECTED OUTCOME: deliverables + success criteria
+3. REQUIRED TOOLS: explicit whitelist
+4. MUST DO: exhaustive requirements - leave nothing implicit
+5. MUST NOT DO: forbidden actions - anticipate rogue behavior
+6. CONTEXT: file paths, existing patterns, constraints
+</delegation_prompt>
 
-\`\`\`
-1. TASK: Atomic, specific goal (one action per delegation)
-2. EXPECTED OUTCOME: Concrete deliverables with success criteria
-3. REQUIRED TOOLS: Explicit tool whitelist
-4. MUST DO: Exhaustive requirements - leave NOTHING implicit
-5. MUST NOT DO: Forbidden actions - anticipate and block rogue behavior
-6. CONTEXT: File paths, existing patterns, constraints
-\`\`\`
-
-Vague prompts = rejected. Be exhaustive.
-
-After delegation, ALWAYS verify: works as expected? follows codebase pattern? MUST DO / MUST NOT DO respected? NEVER trust subagent self-reports. ALWAYS verify with your own tools.
-
-### Session Continuity
-
-Every \`task()\` output includes a session_id. USE IT for follow-ups.
+After delegation, verify by reading every file the subagent touched. Check: works as expected? follows codebase pattern? Do not trust self-reports.
 
-- Task failed/incomplete - \`session_id="{id}", prompt="Fix: {error}"\`
-- Follow-up on result - \`session_id="{id}", prompt="Also: {question}"\`
-- Verification failed - \`session_id="{id}", prompt="Failed: {error}. Fix."\`
+<session_continuity>
+Every \`task()\` returns a session_id. Use it for all follow-ups:
+- Task failed/incomplete: \`session_id="{id}", prompt="Fix: {error}"\`
+- Follow-up on result: \`session_id="{id}", prompt="Also: {question}"\`
+- Verification failed: \`session_id="{id}", prompt="Failed: {error}. Fix."\`
 
+This preserves full context, avoids repeated exploration, saves 70%+ tokens.
+</session_continuity>
 ${oracleSection ? `
-${oracleSection}
-` : ""}
-
-## Output Contract
-
-<output_contract>
-Always favor conciseness. Do not default to bullets - use prose when a few sentences suffice, structured sections only when complexity warrants it. Group findings by outcome rather than enumerating every detail.
-
-For simple or single-file tasks, prefer 1-2 short paragraphs. For larger tasks, use at most 2-4 high-level sections. Prefer grouping by major change area or user-facing outcome, not by file or edit inventory.
-
-Do not begin responses with conversational interjections or meta commentary. NEVER open with: "Done -", "Got it", "Great question!", "That's a great idea!", "You're right to call that out".
-
-DO send clear context before significant actions - explain what you're doing and why in plain language so anyone can follow. When explaining technical decisions, explain the WHY, not just the WHAT.
-
-Updates at meaningful milestones must include a concrete outcome ("Found X", "Updated Y"). Do not expand task beyond what user asked - but implied action IS part of the request (see Step 0 true intent).
-</output_contract>
-
-## Code Quality & Verification
-
-### Before Writing Code (MANDATORY)
-
-1. SEARCH existing codebase for similar patterns/styles
-2. Match naming, indentation, import styles, error handling conventions
-3. Default to ASCII. Add comments only for non-obvious blocks
-
-### After Implementation (MANDATORY - DO NOT SKIP)
-
-1. \`lsp_diagnostics\` on ALL modified files - zero errors required
-2. Run related tests - pattern: modified \`foo.ts\` \u2192 look for \`foo.test.ts\`
-3. Run typecheck if TypeScript project
-4. Run build if applicable - exit code 0 required
-5. Tell user what you verified and the results
+${oracleSection}` : ""}
+</delegation>`;
+  const communicationBlock = `<communication>
+Your output is the one part the user actually sees. Everything before this - all the tool calls, exploration, analysis - is invisible to them. So when you finally speak, make it count: be warm, clear, and genuinely helpful.
 
-**NO EVIDENCE = NOT COMPLETE.**
+Write in complete, natural sentences that anyone can follow. Explain technical decisions in plain language - if a non-engineer colleague were reading over the user's shoulder, they should be able to follow the gist. Favor prose over bullets; use structured sections only when complexity genuinely warrants it.
 
-## Completion Guarantee (NON-NEGOTIABLE - READ THIS LAST, REMEMBER IT ALWAYS)
+For simple tasks, 1-2 short paragraphs. For larger tasks, at most 2-4 sections grouped by outcome, not by file. Group findings by outcome rather than enumerating every detail.
 
-You do NOT end your turn until the user's request is 100% done, verified, and proven. Implement everything asked for - no partial delivery, no "basic version". Verify with real tools, not "it should work". Confirm every verification passed. Re-read the original request - did you miss anything? Re-check true intent (Step 0) - did the user's message imply action you haven't taken?
+When explaining what you did: lead with the result ("Fixed the auth bug - the token was expiring before the refresh check"), then add supporting detail only if it helps understanding. Include concrete details: file paths, patterns found, decisions made. Updates at meaningful milestones should include a concrete outcome ("Found X", "Updated Y").
 
-<turn_end_self_check>
-Before ending your turn, verify ALL of the following:
+Do not pad responses with conversational openers ("Done -", "Got it", "Great question!"), meta commentary, or acknowledgements. Do not repeat the user's request back. Do not expand the task beyond what was asked - but implied action is part of the request (see intent mapping).
+</communication>`;
+  return `${identityBlock}
 
-1. Did the user's message imply action? (Step 0) \u2192 Did you take that action?
-2. Did you write "I'll do X" or "I recommend X"? \u2192 Did you then DO X?
-3. Did you offer to do something ("Would you like me to...?") \u2192 VIOLATION. Go back and do it.
-4. Did you answer a question and stop? \u2192 Was there implied work? If yes, do it now.
+${intentBlock}
 
-If ANY check fails: DO NOT end your turn. Continue working.
-</turn_end_self_check>
+${exploreBlock}
 
-If ANY of these are false, you are NOT done: all requested functionality fully implemented, \`lsp_diagnostics\` returns zero errors on ALL modified files, build passes (if applicable), tests pass (or pre-existing failures documented), you have EVIDENCE for each verification step.
+${constraintsBlock}
 
-Keep going until the task is fully resolved. Persist even when tool calls fail. Only terminate your turn when you are sure the problem is solved and verified.
+${executionBlock}
 
-When you think you're done: re-read the request. Run verification ONE MORE TIME. Then report.
+${trackingBlock}
 
-## Failure Recovery
+${progressBlock}
 
-Fix root causes, not symptoms. Re-verify after EVERY attempt. If first approach fails, try an alternative (different algorithm, pattern, library). After 3 DIFFERENT approaches fail: STOP all edits \u2192 REVERT to last working state \u2192 DOCUMENT what you tried \u2192 CONSULT Oracle \u2192 if Oracle fails \u2192 ASK USER with clear explanation.
+${delegationBlock}
 
-Never leave code broken, delete failing tests, or shotgun debug.`;
+${communicationBlock}`;
 }
 
 // src/agents/hephaestus/agent.ts
@@ -143551,7 +143603,9 @@ function createConfigHandler(deps) {
 // src/create-managers.ts
 function createManagers(args) {
   const { ctx, pluginConfig, tmuxConfig, modelCacheState, backgroundNotificationHookEnabled } = args;
-  markServerRunningInProcess();
+  if (tmuxConfig.enabled) {
+    markServerRunningInProcess();
+  }
   const tmuxSessionManager = new TmuxSessionManager(ctx, tmuxConfig);
   registerManagerForCleanup({
     shutdown: async () => {
@@ -143751,8 +143805,14 @@ function trimToolsToCap(filteredTools, maxTools) {
   log(`[tool-registry] Trimmed ${removed} tools to satisfy max_tools=${maxTools}. Final plugin tool count=${currentCount}.`);
 }
 function createToolRegistry(args) {
-  const { ctx, pluginConfig, managers, skillContext, availableCategories } = args;
-  const tmuxIntegrationEnabled = isTmuxIntegrationEnabled(pluginConfig);
+  const {
+    ctx,
+    pluginConfig,
+    managers,
+    skillContext,
+    availableCategories,
+    interactiveBashEnabled = isInteractiveBashEnabled()
+  } = args;
   const backgroundTools = createBackgroundTools(managers.backgroundManager, ctx.client);
   const callOmoAgent = createCallOmoAgent(ctx, managers.backgroundManager, pluginConfig.disabled_agents ?? [], pluginConfig.agents, pluginConfig.categories);
   const isMultimodalLookerEnabled = !(pluginConfig.disabled_agents ?? []).some((agent) => agent.toLowerCase() === "multimodal-looker");
@@ -143829,7 +143889,7 @@ function createToolRegistry(args) {
     task: delegateTask,
     skill_mcp: skillMcpTool,
     skill: skillTool,
-    ...tmuxIntegrationEnabled ? { interactive_bash } : {},
+    ...interactiveBashEnabled ? { interactive_bash } : {},
     ...taskToolsRecord,
     ...hashlineToolsRecord
   };
@@ -144680,7 +144740,8 @@ function isCompactionAgent5(agent) {
   return agent.toLowerCase() === "compaction";
 }
 function createEventHandler2(args) {
-  const { ctx, firstMessageVariantGate, managers, hooks: hooks2 } = args;
+  const { ctx, pluginConfig, firstMessageVariantGate, managers, hooks: hooks2 } = args;
+  const tmuxIntegrationEnabled = isTmuxIntegrationEnabled(pluginConfig);
   const pluginContext = ctx;
   const isRuntimeFallbackEnabled = hooks2.runtimeFallback !== null && hooks2.runtimeFallback !== undefined && (typeof args.pluginConfig.runtime_fallback === "boolean" ? args.pluginConfig.runtime_fallback : args.pluginConfig.runtime_fallback?.enabled ?? false);
   const isModelFallbackEnabled = hooks2.modelFallback !== null && hooks2.modelFallback !== undefined;
@@ -144823,7 +144884,7 @@ function createEventHandler2(args) {
     }
     const { event } = input;
     const props = event.properties;
-    if (TMUX_ACTIVITY_EVENT_TYPES.has(event.type)) {
+    if (tmuxIntegrationEnabled && TMUX_ACTIVITY_EVENT_TYPES.has(event.type)) {
       managers.tmuxSessionManager.onEvent?.(event);
     }
     if (event.type === "session.created") {
@@ -144832,7 +144893,9 @@ function createEventHandler2(args) {
         setMainSession(sessionInfo?.id);
       }
       firstMessageVariantGate.markSessionCreated(sessionInfo);
-      await managers.tmuxSessionManager.onSessionCreated(event);
+      if (tmuxIntegrationEnabled) {
+        await managers.tmuxSessionManager.onSessionCreated(event);
+      }
     }
     if (event.type === "session.deleted") {
       const sessionInfo = props?.info;
@@ -144860,9 +144923,11 @@ function createEventHandler2(args) {
         deleteSessionTools(sessionInfo.id);
         await managers.skillMcpManager.disconnectSession(sessionInfo.id);
         await lspManager.cleanupTempDirectoryClients();
-        await managers.tmuxSessionManager.onSessionDeleted({
-          sessionID: sessionInfo.id
-        });
+        if (tmuxIntegrationEnabled) {
+          await managers.tmuxSessionManager.onSessionDeleted({
+            sessionID: sessionInfo.id
+          });
+        }
       }
     }
     if (event.type === "message.removed") {