claudecode-omc 4.8.4 → 4.9.4

package/scripts/post-tool-verifier.mjs

@@ -14,6 +14,13 @@ import { readStdin } from './lib/stdin.mjs';
 
 const AGENT_OUTPUT_ANALYSIS_LIMIT = parseInt(process.env.OMC_AGENT_OUTPUT_ANALYSIS_LIMIT || '12000', 10);
 const AGENT_OUTPUT_SUMMARY_LIMIT = parseInt(process.env.OMC_AGENT_OUTPUT_SUMMARY_LIMIT || '360', 10);
+const QUIET_LEVEL = getQuietLevel();
+
+function getQuietLevel() {
+  const parsed = Number.parseInt(process.env.OMC_QUIET || '0', 10);
+  if (Number.isNaN(parsed)) return 0;
+  return Math.max(0, parsed);
+}
 
 // Get the directory of this script to resolve the dist module
 const __filename = fileURLToPath(import.meta.url);
@@ -138,7 +145,8 @@ function stripClaudeTempCwdErrors(output) {
 }
 
 // Pattern matching Claude Code's "Error: Exit code N" prefix line
-const CLAUDE_EXIT_CODE_PREFIX = /^Error: Exit code \d+\s*$/gm;
+// Note: no /g flag — module-level regex with /g is stateful (.lastIndex persists across calls)
+const CLAUDE_EXIT_CODE_PREFIX = /^Error: Exit code \d+\s*$/m;
 
 /**
  * Detect non-zero exit code with valid stdout (issue #960).
@@ -152,12 +160,9 @@ export function isNonZeroExitWithOutput(output) {
 
   // Must contain Claude Code's exit code prefix
   if (!CLAUDE_EXIT_CODE_PREFIX.test(cleaned)) return false;
-  // Reset regex state (global flag)
-  CLAUDE_EXIT_CODE_PREFIX.lastIndex = 0;
 
   // Strip exit code prefix line(s) and check remaining content
   const remaining = cleaned.replace(CLAUDE_EXIT_CODE_PREFIX, '').trim();
-  CLAUDE_EXIT_CODE_PREFIX.lastIndex = 0;
 
   // Must have at least one non-empty line of real output
   const contentLines = remaining.split('\n').filter(l => l.trim().length > 0);
@@ -299,7 +304,7 @@ export function detectWriteFailure(output) {
 }
 
 // Get agent completion summary from tracking state
-function getAgentCompletionSummary(directory) {
+function getAgentCompletionSummary(directory, quietLevel = QUIET_LEVEL) {
   const trackingFile = join(directory, '.omc', 'state', 'subagent-tracking.json');
   try {
     if (existsSync(trackingFile)) {
@@ -312,10 +317,10 @@ function getAgentCompletionSummary(directory) {
       if (running.length === 0 && completed === 0 && failed === 0) return '';
 
       const parts = [];
-      if (running.length > 0) {
+      if (quietLevel < 2 && running.length > 0) {
         parts.push(`Running: ${running.length} [${running.map(a => a.agent_type.replace('oh-my-claudecode:', '')).join(', ')}]`);
       }
-      if (completed > 0) parts.push(`Completed: ${completed}`);
+      if (quietLevel < 2 && completed > 0) parts.push(`Completed: ${completed}`);
       if (failed > 0) parts.push(`Failed: ${failed}`);
 
       return parts.join(' | ');
@@ -338,7 +343,7 @@ function generateMessage(toolName, toolOutput, sessionId, toolCount, directory,
         message = `Command exited with code ${code} but produced valid output. This may be expected behavior.`;
       } else if (detectBashFailure(toolOutput)) {
         message = 'Command failed. Please investigate the error and fix before continuing.';
-      } else if (detectBackgroundOperation(toolOutput)) {
+      } else if (QUIET_LEVEL < 2 && detectBackgroundOperation(toolOutput)) {
         message = 'Background operation detected. Remember to verify results before proceeding.';
       }
       break;
@@ -346,12 +351,12 @@ function generateMessage(toolName, toolOutput, sessionId, toolCount, directory,
     case 'Task':
     case 'TaskCreate':
     case 'TaskUpdate': {
-      const agentSummary = getAgentCompletionSummary(directory);
+      const agentSummary = getAgentCompletionSummary(directory, QUIET_LEVEL);
       if (detectWriteFailure(toolOutput)) {
         message = 'Task delegation failed. Verify agent name and parameters.';
-      } else if (detectBackgroundOperation(toolOutput)) {
+      } else if (QUIET_LEVEL < 2 && detectBackgroundOperation(toolOutput)) {
         message = 'Background task launched. Use TaskOutput to check results when needed.';
-      } else if (toolCount > 5) {
+      } else if (QUIET_LEVEL < 2 && toolCount > 5) {
         message = `Multiple tasks delegated (${toolCount} total). Track their completion status.`;
       }
       if (wasTruncated) {
@@ -366,7 +371,7 @@ function generateMessage(toolName, toolOutput, sessionId, toolCount, directory,
 
     case 'TaskOutput': {
       const summary = summarizeAgentResult(toolOutput);
-      if (summary) {
+      if (QUIET_LEVEL < 2 && summary) {
         message = `TaskOutput summary: ${summary}`;
       }
       if (wasTruncated) {
@@ -379,7 +384,7 @@ function generateMessage(toolName, toolOutput, sessionId, toolCount, directory,
     case 'Edit':
       if (detectWriteFailure(toolOutput)) {
         message = 'Edit operation failed. Verify file exists and content matches exactly.';
-      } else {
+      } else if (QUIET_LEVEL === 0) {
         message = 'Code modified. Verify changes work as expected before marking complete.';
       }
       break;
@@ -387,35 +392,35 @@ function generateMessage(toolName, toolOutput, sessionId, toolCount, directory,
     case 'Write':
       if (detectWriteFailure(toolOutput)) {
         message = 'Write operation failed. Check file permissions and directory existence.';
-      } else {
+      } else if (QUIET_LEVEL === 0) {
         message = 'File written. Test the changes to ensure they work correctly.';
       }
       break;
 
     case 'TodoWrite':
-      if (/created|added/i.test(toolOutput)) {
+      if (QUIET_LEVEL === 0 && /created|added/i.test(toolOutput)) {
         message = 'Todo list updated. Proceed with next task on the list.';
-      } else if (/completed|done/i.test(toolOutput)) {
+      } else if (QUIET_LEVEL === 0 && /completed|done/i.test(toolOutput)) {
         message = 'Task marked complete. Continue with remaining todos.';
-      } else if (/in_progress/i.test(toolOutput)) {
+      } else if (QUIET_LEVEL === 0 && /in_progress/i.test(toolOutput)) {
         message = 'Task marked in progress. Focus on completing this task.';
       }
       break;
 
     case 'Read':
-      if (toolCount > 10) {
+      if (QUIET_LEVEL === 0 && toolCount > 10) {
         message = `Extensive reading (${toolCount} files). Consider using Grep for pattern searches.`;
       }
       break;
 
     case 'Grep':
-      if (/^0$|no matches/i.test(toolOutput)) {
+      if (QUIET_LEVEL === 0 && /^0$|no matches/i.test(toolOutput)) {
         message = 'No matches found. Verify pattern syntax or try broader search.';
       }
       break;
 
     case 'Glob':
-      if (!toolOutput.trim() || /no files/i.test(toolOutput)) {
+      if (QUIET_LEVEL === 0 && (!toolOutput.trim() || /no files/i.test(toolOutput))) {
         message = 'No files matched pattern. Verify glob syntax and directory.';
       }
       break;

package/scripts/pre-tool-enforcer.mjs

@@ -10,9 +10,24 @@ import { closeSync, existsSync, mkdirSync, openSync, readFileSync, readSync, ren
 import { dirname, join, resolve } from 'path';
 import { execSync } from 'child_process';
 import { homedir } from 'os';
-import { pathToFileURL } from 'url';
+import { fileURLToPath, pathToFileURL } from 'url';
 import { readStdin } from './lib/stdin.mjs';
 
+// Inlined from src/config/models.ts — avoids a dist/ import so the hook works
+// before a build and stays consistent with the TypeScript source.
+function isProviderSpecificModelId(modelId) {
+  if (/^((us|eu|ap|global)\.anthropic\.|anthropic\.claude)/i.test(modelId)) return true;
+  if (/^arn:aws(-[^:]+)?:bedrock:/i.test(modelId)) return true;
+  if (modelId.toLowerCase().startsWith('vertex_ai/')) return true;
+  return false;
+}
+function hasExtendedContextSuffix(modelId) {
+  return /\[\d+[mk]\]$/i.test(modelId);
+}
+function isSubagentSafeModelId(modelId) {
+  return isProviderSpecificModelId(modelId) && !hasExtendedContextSuffix(modelId);
+}
+
 const SESSION_ID_PATTERN = /^[a-zA-Z0-9][a-zA-Z0-9_-]{0,255}$/;
 const MODE_STATE_FILES = [
   'autopilot-state.json',
@@ -26,6 +41,13 @@ const MODE_STATE_FILES = [
 ];
 const AGENT_HEAVY_TOOLS = new Set(['Task', 'TaskCreate', 'TaskUpdate']);
 const PREFLIGHT_CONTEXT_THRESHOLD = parseInt(process.env.OMC_AGENT_PREFLIGHT_CONTEXT_THRESHOLD || '72', 10);
+const QUIET_LEVEL = getQuietLevel();
+
+function getQuietLevel() {
+  const parsed = Number.parseInt(process.env.OMC_QUIET || '0', 10);
+  if (Number.isNaN(parsed)) return 0;
+  return Math.max(0, parsed);
+}
 
 /**
  * Resolve transcript path in worktree environments.
@@ -268,6 +290,7 @@ function getActiveTeamState(directory, sessionId) {
 // Generate agent spawn message with metadata
 function generateAgentSpawnMessage(toolInput, directory, todoStatus, sessionId) {
   if (!toolInput || typeof toolInput !== 'object') {
+    if (QUIET_LEVEL >= 2) return '';
     return `${todoStatus}Launch multiple agents in parallel when tasks are independent. Use run_in_background for long operations.`;
   }
 
@@ -291,6 +314,8 @@ function generateAgentSpawnMessage(toolInput, directory, todoStatus, sessionId)
       `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 is set in ~/.claude/settings.json and restart Claude Code.`;
   }
 
+  if (QUIET_LEVEL >= 2) return '';
+
   const parts = [`${todoStatus}Spawning agent: ${agentType} (${model})${bg}`];
   if (desc) parts.push(`Task: ${desc}`);
   if (tracking.running > 0) parts.push(`Active agents: ${tracking.running}`);
@@ -300,6 +325,13 @@ function generateAgentSpawnMessage(toolInput, directory, todoStatus, sessionId)
 
 // Generate contextual message based on tool type
 function generateMessage(toolName, todoStatus, modeActive = false) {
+  if (QUIET_LEVEL >= 1 && ['Bash', 'Edit', 'Write', 'Read', 'Grep', 'Glob'].includes(toolName)) {
+    return '';
+  }
+  if (QUIET_LEVEL >= 2 && toolName === 'TodoWrite') {
+    return '';
+  }
+
   const messages = {
     TodoWrite: `${todoStatus}Mark todos in_progress BEFORE starting, completed IMMEDIATELY after finishing.`,
     Bash: `${todoStatus}Use parallel execution for independent tasks. Use run_in_background for long operations (npm install, builds, tests).`,
@@ -343,9 +375,39 @@ const SKILL_PROTECTION_MAP = {
   deepinit: 'heavy',
 };
 
-function getSkillProtectionLevel(skillName) {
+function getSkillProtectionLevel(skillName, rawSkillName) {
+  // When rawSkillName is provided, only apply protection to OMC-prefixed skills.
+  // Non-prefixed skills are project custom skills or other plugins — no protection.
+  // See: https://github.com/Yeachan-Heo/oh-my-claudecode/issues/1581
+  if (rawSkillName != null && typeof rawSkillName === 'string' &&
+      !rawSkillName.toLowerCase().startsWith('oh-my-claudecode:')) {
+    return 'none';
+  }
   const normalized = (skillName || '').toLowerCase().replace(/^oh-my-claudecode:/, '');
-  return SKILL_PROTECTION_MAP[normalized] || 'light';
+  return SKILL_PROTECTION_MAP[normalized] || 'none';
+}
+
+// Load OMC config to check forceInherit setting (issues #1135, #1201)
+function loadOmcConfig() {
+  const configPaths = [
+    join(homedir(), '.claude', '.omc-config.json'),
+    join(process.cwd(), '.omc', 'config.json'),
+  ];
+  for (const configPath of configPaths) {
+    try {
+      if (existsSync(configPath)) {
+        return JSON.parse(readFileSync(configPath, 'utf-8'));
+      }
+    } catch { /* continue */ }
+  }
+  return {};
+}
+
+// Check if forceInherit is enabled via config or env var
+function isForceInheritEnabled() {
+  if (process.env.OMC_ROUTING_FORCE_INHERIT === 'true') return true;
+  const config = loadOmcConfig();
+  return config.routing?.forceInherit === true;
 }
 
 function extractSkillName(toolInput) {
@@ -356,8 +418,8 @@ function extractSkillName(toolInput) {
   return normalized.includes(':') ? normalized.split(':').at(-1).toLowerCase() : normalized.toLowerCase();
 }
 
-function writeSkillActiveState(directory, skillName, sessionId) {
-  const protection = getSkillProtectionLevel(skillName);
+function writeSkillActiveState(directory, skillName, sessionId, rawSkillName) {
+  const protection = getSkillProtectionLevel(skillName, rawSkillName);
   if (protection === 'none') return;
 
   const config = SKILL_PROTECTION_CONFIGS[protection];
@@ -394,6 +456,50 @@ function writeSkillActiveState(directory, skillName, sessionId) {
   }
 }
 
+
+function clearAwaitingConfirmationFlag(directory, stateName, sessionId) {
+  const stateDir = join(directory, '.omc', 'state');
+  const safeSessionId = sessionId && SESSION_ID_PATTERN.test(sessionId) ? sessionId : '';
+  const paths = [
+    safeSessionId ? join(stateDir, 'sessions', safeSessionId, `${stateName}-state.json`) : null,
+    join(stateDir, `${stateName}-state.json`),
+  ].filter(Boolean);
+
+  for (const statePath of paths) {
+    try {
+      if (!existsSync(statePath)) continue;
+      const state = JSON.parse(readFileSync(statePath, 'utf-8'));
+      if (!state || typeof state !== 'object' || !state.awaiting_confirmation) continue;
+      delete state.awaiting_confirmation;
+      const tmpPath = statePath + '.tmp';
+      writeFileSync(tmpPath, JSON.stringify(state, null, 2), { mode: 0o600 });
+      renameSync(tmpPath, statePath);
+    } catch {
+      // Best-effort; don't fail the hook
+    }
+  }
+}
+
+function confirmSkillModeStates(directory, skillName, sessionId) {
+  switch (skillName) {
+    case 'ralph':
+      clearAwaitingConfirmationFlag(directory, 'ralph', sessionId);
+      clearAwaitingConfirmationFlag(directory, 'ultrawork', sessionId);
+      break;
+    case 'ultrawork':
+      clearAwaitingConfirmationFlag(directory, 'ultrawork', sessionId);
+      break;
+    case 'autopilot':
+      clearAwaitingConfirmationFlag(directory, 'autopilot', sessionId);
+      break;
+    case 'ralplan':
+      clearAwaitingConfirmationFlag(directory, 'ralplan', sessionId);
+      break;
+    default:
+      break;
+  }
+}
+
 // Record Skill/Task invocations to flow trace (best-effort)
 async function recordToolInvocation(data, directory) {
   try {
@@ -439,7 +545,11 @@ async function main() {
       if (skillName) {
         const sid = typeof data.session_id === 'string' ? data.session_id
           : typeof data.sessionId === 'string' ? data.sessionId : '';
-        writeSkillActiveState(directory, skillName, sid);
+        // Pass rawSkillName to distinguish OMC skills from project custom skills (issue #1581)
+        const rawSkill = toolInput.skill || toolInput.skill_name || toolInput.skillName || toolInput.command || '';
+        const rawSkillName = typeof rawSkill === 'string' && rawSkill.trim() ? rawSkill.trim() : undefined;
+        writeSkillActiveState(directory, skillName, sid, rawSkillName);
+        confirmSkillModeStates(directory, skillName, sid);
       }
     }
 
@@ -451,6 +561,74 @@ async function main() {
           : '';
     const modeActive = hasActiveMode(directory, sessionId);
 
+    // Force-inherit check: deny Task/Agent calls with invalid model param when forceInherit is
+    // enabled (Bedrock, Vertex, CC Switch, etc.) - issues #1135, #1201, #1767, #1868
+    //
+    // New behaviour (issue #1868 — [1m] suffix deadlock):
+    //   ALLOW explicit valid provider-specific model IDs (full Bedrock/Vertex format, no [1m])
+    //   DENY  tier names (sonnet/opus/haiku) and [1m]-suffixed IDs
+    //   DENY  no-model calls when the session model itself has [1m] — guide to OMC_SUBAGENT_MODEL
+    if (toolName === 'Task' || toolName === 'Agent') {
+      const toolInput = data.toolInput || data.tool_input || {};
+      const toolModel = toolInput.model;
+      if (isForceInheritEnabled()) {
+        // Check both vars: if either carries [1m] the session model is unsafe for sub-agents.
+        // Avoids a split-brain between the hook and runtime code that may read the vars in
+        // different orders (e.g. model-contract.ts uses ANTHROPIC_MODEL first).
+        const claudeModel = process.env.CLAUDE_MODEL || '';
+        const anthropicModel = process.env.ANTHROPIC_MODEL || '';
+        const sessionHasLmSuffix =
+          hasExtendedContextSuffix(claudeModel) || hasExtendedContextSuffix(anthropicModel);
+        // For error messages: prefer whichever var actually carries the [1m] suffix.
+        const sessionModel = hasExtendedContextSuffix(claudeModel)
+          ? claudeModel
+          : hasExtendedContextSuffix(anthropicModel)
+            ? anthropicModel
+            : claudeModel || anthropicModel;
+
+        if (toolModel) {
+          // Allow explicit valid provider-specific IDs (full Bedrock/Vertex format) without a
+          // [1m] suffix — blocking these leaves no escape hatch when the inherited session model
+          // is itself invalid. Reject tier names (sonnet/opus/haiku) and [1m]-suffixed IDs.
+          if (!isSubagentSafeModelId(toolModel)) {
+            const subagentModel = process.env.OMC_SUBAGENT_MODEL || '';
+            const guidance = subagentModel
+              ? `Pass model="${subagentModel}" (your configured OMC_SUBAGENT_MODEL value).`
+              : `Remove the \`model\` parameter, or set OMC_SUBAGENT_MODEL=<valid-bedrock-id> and pass that value explicitly.`;
+            console.log(JSON.stringify({
+              continue: true,
+              hookSpecificOutput: {
+                hookEventName: 'PreToolUse',
+                permissionDecision: 'deny',
+                permissionDecisionReason: `[MODEL ROUTING] This environment uses a non-standard provider (Bedrock/Vertex/proxy). ${guidance} The model "${toolModel}" is not valid for this provider.`
+              }
+            }));
+            return;
+          }
+          // else: valid provider-specific model ID — fall through to continue.
+        } else if (sessionHasLmSuffix) {
+          // No model param, but the session model has a [1m] context-window suffix.
+          // Sub-agents would inherit it and fail — the runtime strips [1m] to a bare
+          // Anthropic model ID (e.g. claude-sonnet-4-6) which is invalid on Bedrock.
+          const subagentModel = process.env.OMC_SUBAGENT_MODEL || '';
+          const suggestion = subagentModel
+            ? `Pass model="${subagentModel}" (your configured OMC_SUBAGENT_MODEL) explicitly on this ${toolName} call.`
+            : `Set OMC_SUBAGENT_MODEL=<valid-bedrock-id> in your environment (use the model ID from the 400 error message, e.g. "us.anthropic.claude-sonnet-4-5-20250929-v1:0"), then pass that value as the model parameter.`;
+          console.log(JSON.stringify({
+            continue: true,
+            hookSpecificOutput: {
+              hookEventName: 'PreToolUse',
+              permissionDecision: 'deny',
+              permissionDecisionReason: `[MODEL ROUTING] Your session model "${sessionModel}" has a context-window suffix ([1m]) that sub-agents cannot inherit — the runtime strips it to a bare Anthropic model ID which is invalid on Bedrock. ${suggestion}`
+            }
+          }));
+          return;
+        }
+        // else: no model param and no [1m] on session model → normal forceInherit,
+        // agents inherit the parent session's model cleanly.
+      }
+    }
+
     // Send notification when AskUserQuestion is about to execute (user input needed)
     // Fires in PreToolUse so users get notified BEFORE the tool blocks for input (#597)
     if (toolName === 'AskUserQuestion') {

package/scripts/project-memory-session.mjs

@@ -11,6 +11,10 @@ import { fileURLToPath, pathToFileURL } from 'url';
 const __filename = fileURLToPath(import.meta.url);
 const __dirname = dirname(__filename);
 
+function getRuntimeBaseDir() {
+  return process.env.CLAUDE_PLUGIN_ROOT || join(__dirname, '..');
+}
+
 // Import timeout-protected stdin reader (prevents hangs on Linux/Windows, see issue #240, #524)
 let readStdin;
 try {
@@ -34,7 +38,7 @@ try {
 // Dynamic import of project memory module (prevents crash if dist is missing, see issue #362)
 let registerProjectMemoryContext;
 try {
-  const mod = await import(pathToFileURL(join(__dirname, '..', 'dist', 'hooks', 'project-memory', 'index.js')).href);
+  const mod = await import(pathToFileURL(join(getRuntimeBaseDir(), 'dist', 'hooks', 'project-memory', 'index.js')).href);
   registerProjectMemoryContext = mod.registerProjectMemoryContext;
 } catch {
   // dist not built or missing - skip project memory detection silently