oh-my-opencode 4.16.3 → 4.17.0

package/dist/index.js

@@ -21139,6 +21139,11 @@ var CATEGORY_MODEL_REQUIREMENTS = {
   },
   deep: {
     fallbackChain: [
+      {
+        providers: ["openai", "vercel"],
+        model: "gpt-5.6-terra",
+        variant: "xhigh"
+      },
       {
         providers: ["openai", "vercel"],
         model: "gpt-5.6-sol",
@@ -21202,6 +21207,11 @@ var CATEGORY_MODEL_REQUIREMENTS = {
   },
   "unspecified-low": {
     fallbackChain: [
+      {
+        providers: ["openai", "vercel"],
+        model: "gpt-5.6-luna",
+        variant: "xhigh"
+      },
       {
         providers: ["anthropic", "github-copilot", "opencode", "vercel"],
         model: "claude-sonnet-4-6"
@@ -22603,6 +22613,38 @@ var SUPPLEMENTAL_MODEL_CAPABILITIES = {
       output: 128000
     }
   },
+  "gpt-5.6-terra": {
+    id: "gpt-5.6-terra",
+    family: "gpt-mini",
+    reasoning: true,
+    temperature: false,
+    toolCall: true,
+    modalities: {
+      input: ["text", "image", "pdf"],
+      output: ["text"]
+    },
+    limit: {
+      context: 1050000,
+      input: 922000,
+      output: 128000
+    }
+  },
+  "gpt-5.6-luna": {
+    id: "gpt-5.6-luna",
+    family: "gpt-nano",
+    reasoning: true,
+    temperature: false,
+    toolCall: true,
+    modalities: {
+      input: ["text", "image", "pdf"],
+      output: ["text"]
+    },
+    limit: {
+      context: 1050000,
+      input: 922000,
+      output: 128000
+    }
+  },
   "gpt-5.5": {
     id: "gpt-5.5",
     family: "gpt",
@@ -23506,28 +23548,6 @@ function truncateDescription(description, maxLength = 120) {
 }
 
 // packages/omo-opencode/src/tools/delegate-task/anthropic-categories.ts
-var UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND = `<Category_Context>
-You are working on tasks that don't fit specific categories but require moderate effort.
-
-<Selection_Gate>
-BEFORE selecting this category, VERIFY ALL conditions:
-1. Task does NOT fit: quick (trivial), visual-engineering (UI), ultrabrain (deep logic), artistry (creative), writing (docs)
-2. Task requires more than trivial effort but is NOT system-wide
-3. Scope is contained within a few files/modules
-
-If task fits ANY other category, DO NOT select unspecified-low.
-This is NOT a default choice - it's for genuinely unclassifiable moderate-effort work.
-</Selection_Gate>
-</Category_Context>
-
-<Caller_Warning>
-THIS CATEGORY USES A MID-TIER MODEL (claude-sonnet-4-6).
-
-**PROVIDE CLEAR STRUCTURE:**
-1. MUST DO: Enumerate required actions explicitly
-2. MUST NOT DO: State forbidden actions to prevent scope creep
-3. EXPECTED OUTPUT: Define concrete success criteria
-</Caller_Warning>`;
 var UNSPECIFIED_HIGH_CATEGORY_PROMPT_APPEND = `<Category_Context>
 You are working on tasks that don't fit specific categories but require substantial effort.
 
@@ -23543,12 +23563,6 @@ If task is unclassifiable but moderate-effort, use unspecified-low instead.
 </Selection_Gate>
 </Category_Context>`;
 var ANTHROPIC_CATEGORIES = [
-  {
-    name: "unspecified-low",
-    config: { model: "anthropic/claude-sonnet-4-6" },
-    description: "Tasks that don't fit other categories, low effort required",
-    promptAppend: UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND
-  },
   {
     name: "unspecified-high",
     config: { model: "anthropic/claude-opus-4-7", variant: "max" },
@@ -23856,6 +23870,28 @@ EXPECTED OUTPUT:
 
 If your prompt lacks this structure, REWRITE IT before delegating.
 </Caller_Warning>`;
+var UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND = `<Category_Context>
+You are working on tasks that don't fit specific categories but require moderate effort.
+
+<Selection_Gate>
+BEFORE selecting this category, VERIFY ALL conditions:
+1. Task does NOT fit: quick (trivial), visual-engineering (UI), ultrabrain (deep logic), artistry (creative), writing (docs)
+2. Task requires more than trivial effort but is NOT system-wide
+3. Scope is contained within a few files/modules
+
+If task fits ANY other category, DO NOT select unspecified-low.
+This is NOT a default choice - it's for genuinely unclassifiable moderate-effort work.
+</Selection_Gate>
+</Category_Context>
+
+<Caller_Warning>
+THIS CATEGORY USES A LIGHTWEIGHT MODEL (gpt-5.6-luna).
+
+**PROVIDE CLEAR STRUCTURE:**
+1. MUST DO: Enumerate required actions explicitly
+2. MUST NOT DO: State forbidden actions to prevent scope creep
+3. EXPECTED OUTPUT: Define concrete success criteria
+</Caller_Warning>`;
 var OPENAI_CATEGORIES = [
   {
     name: "ultrabrain",
@@ -23865,7 +23901,7 @@ var OPENAI_CATEGORIES = [
   },
   {
     name: "deep",
-    config: { model: "openai/gpt-5.6-sol", variant: "high" },
+    config: { model: "openai/gpt-5.6-terra", variant: "xhigh" },
     description: "Goal-oriented autonomous problem-solving on hairy problems requiring deep research. ONE goal + ONE deliverable per call \u2014 multiple goals must fan out as parallel `deep` calls, never bundled into one.",
     promptAppend: DEEP_CATEGORY_PROMPT_APPEND,
     resolvePromptAppend: resolveDeepCategoryPromptAppend
@@ -23875,6 +23911,12 @@ var OPENAI_CATEGORIES = [
     config: { model: "openai/gpt-5.4-mini" },
     description: "Trivial tasks - single file changes, typo fixes, simple modifications",
     promptAppend: QUICK_CATEGORY_PROMPT_APPEND
+  },
+  {
+    name: "unspecified-low",
+    config: { model: "openai/gpt-5.6-luna", variant: "xhigh" },
+    description: "Tasks that don't fit other categories, low effort required",
+    promptAppend: UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND
   }
 ];
 
@@ -89613,7 +89655,7 @@ function getCachedVersion(options = {}) {
 // package.json
 var package_default = {
   name: "oh-my-opencode",
-  version: "4.16.3",
+  version: "4.17.0",
   description: "The Best AI Agent Harness - Batteries-Included OpenCode Plugin with Multi-Model Orchestration, Parallel Background Agents, and Crafted LSP/AST Tools",
   main: "./dist/index.js",
   types: "dist/index.d.ts",
@@ -89676,6 +89718,7 @@ var package_default = {
     "packages/lsp-tools-mcp/dist",
     "packages/lsp-daemon/package.json",
     "packages/lsp-daemon/dist",
+    "packages/git-bash-mcp/package.json",
     "packages/git-bash-mcp/dist",
     "packages/shared-skills/package.json",
     "packages/shared-skills/index.mjs",
@@ -89741,7 +89784,7 @@ var package_default = {
     "typecheck:packages": "tsgo --noEmit -p packages/rules-engine/tsconfig.json && tsgo --noEmit -p packages/delegate-core/tsconfig.json && tsgo --noEmit -p packages/mcp-stdio-core/tsconfig.json && tsgo --noEmit -p packages/mcp-client-core/tsconfig.json && tsgo --noEmit -p packages/git-bash-mcp/tsconfig.json && tsgo --noEmit -p packages/lsp-core/tsconfig.json && tsgo --noEmit -p packages/utils/tsconfig.json && tsgo --noEmit -p packages/model-core/tsconfig.json && tsgo --noEmit -p packages/omo-config-core/tsconfig.json && tsgo --noEmit -p packages/prompts-core/tsconfig.json && tsgo --noEmit -p packages/comment-checker-core/tsconfig.json && tsgo --noEmit -p packages/hashline-core/tsconfig.json && tsgo --noEmit -p packages/tmux-core/tsconfig.json && tsgo --noEmit -p packages/team-core/tsconfig.json && tsgo --noEmit -p packages/openclaw-core/tsconfig.json && tsgo --noEmit -p packages/boulder-state/tsconfig.json && tsgo --noEmit -p packages/telemetry-core/tsconfig.json && tsgo --noEmit -p packages/claude-code-compat-core/tsconfig.json && tsgo --noEmit -p packages/skills-loader-core/tsconfig.json && tsgo --noEmit -p packages/agents-md-core/tsconfig.json && tsgo --noEmit -p packages/omo-codex/plugin/shared/tsconfig.json && tsgo --noEmit -p packages/omo-codex/tsconfig.json && tsgo --noEmit -p packages/omo-senpi/tsconfig.json && tsgo --noEmit -p packages/senpi-task/tsconfig.json && tsgo --noEmit -p packages/pi-goal/tsconfig.json && tsgo --noEmit -p packages/pi-webfetch/tsconfig.json && tsgo --noEmit -p packages/omo-opencode/tsconfig.json",
     "typecheck:script": "tsgo --noEmit -p script/tsconfig.json",
     test: "bun test",
-    "test:codex": "bun run build:codex-install && bun run build:git-bash-mcp && bun run build:lsp-tools-mcp && bun run build:lsp-daemon && npm --prefix packages/lsp-tools-mcp test && npm --prefix packages/omo-codex/plugin ci && bun run --cwd packages/omo-codex/plugin build && npm --prefix packages/omo-codex/plugin/components/codegraph run typecheck && npm --prefix packages/omo-codex/plugin/components/codegraph test && node scripts/check-third-party-notices.mjs --ship && bun test packages/omo-opencode/src/cli/cli-installer.platform.test.ts packages/omo-codex/src/install/codex-cache.test.ts packages/omo-codex/src/install/codex-cleanup.test.ts packages/omo-codex/src/install/codex-config-agent-cleanup.test.ts packages/omo-codex/src/install/codex-config-autonomous-features.test.ts packages/omo-codex/src/install/codex-config-reasoning.test.ts packages/omo-codex/src/install/codex-config-toml.test.ts packages/omo-codex/src/install/codex-project-local-cleanup.test.ts packages/omo-codex/src/install/install-codex-project-local-cleanup.test.ts packages/omo-codex/src/install/install-codex.test.ts packages/omo-codex/src/install/install-codex-packaged.test.ts packages/omo-codex/src/install/link-cached-plugin-agents.test.ts packages/omo-codex/src/**/*.test.ts packages/utils/src/jsonc-parser.test.ts packages/utils/src/frontmatter.test.ts packages/hashline-core/src/hash-computation.test.ts packages/hashline-core/src/smoke-untested-modules.test.ts packages/rules-engine/src/index.test.ts packages/rules-engine/src/security-boundary.test.ts packages/agents-md-core/src/injector.test.ts packages/omo-codex/plugin/components/lsp/test/package-smoke.test.ts && node --test packages/omo-codex/plugin/test/*.test.mjs packages/omo-codex/scripts/install-cache-copy.test.mjs packages/omo-codex/scripts/install-cli-args.test.mjs packages/omo-codex/scripts/install-delegated-command.test.mjs packages/omo-codex/scripts/install-config-autonomous-features.test.mjs packages/omo-codex/scripts/install-config-autonomous.test.mjs packages/omo-codex/scripts/install-config-reasoning.test.mjs packages/omo-codex/scripts/install-config.test.mjs packages/omo-codex/scripts/install-hook-targets.test.mjs packages/omo-codex/scripts/install-project-local-cleanup.test.mjs packages/omo-codex/scripts/install-lazycodex-version-stamp.test.mjs packages/omo-codex/scripts/install-local-entrypoint.test.mjs packages/omo-codex/scripts/install-local-git-bash-preflight.test.mjs packages/omo-codex/scripts/install-local.test.mjs packages/omo-codex/scripts/install-marketplace-cache.test.mjs packages/omo-codex/scripts/install-mcp-context7-runtime.test.mjs packages/omo-codex/scripts/install-mcp-runtime.test.mjs packages/omo-codex/scripts/install-packaged-local.test.mjs packages/omo-codex/scripts/install-generated-bundle.test.mjs packages/omo-codex/scripts/install-agent-links.test.mjs packages/omo-codex/scripts/install-bin-links.test.mjs",
+    "test:codex": "bun run build:codex-install && bun run build:git-bash-mcp && bun run build:lsp-tools-mcp && bun run build:lsp-daemon && npm --prefix packages/lsp-tools-mcp test && npm --prefix packages/omo-codex/plugin ci && npm --prefix packages/omo-codex/plugin/components/ulw-loop test && bun run --cwd packages/omo-codex/plugin build && npm --prefix packages/omo-codex/plugin/components/codegraph run typecheck && npm --prefix packages/omo-codex/plugin/components/codegraph test && node scripts/check-third-party-notices.mjs --ship && bun test packages/omo-opencode/src/cli/cli-installer.platform.test.ts packages/omo-codex/src/install/codex-cache.test.ts packages/omo-codex/src/install/codex-cleanup.test.ts packages/omo-codex/src/install/codex-config-agent-cleanup.test.ts packages/omo-codex/src/install/codex-config-autonomous-features.test.ts packages/omo-codex/src/install/codex-config-reasoning.test.ts packages/omo-codex/src/install/codex-config-toml.test.ts packages/omo-codex/src/install/codex-project-local-cleanup.test.ts packages/omo-codex/src/install/install-codex-project-local-cleanup.test.ts packages/omo-codex/src/install/install-codex.test.ts packages/omo-codex/src/install/install-codex-packaged.test.ts packages/omo-codex/src/install/link-cached-plugin-agents.test.ts packages/omo-codex/src/**/*.test.ts packages/utils/src/jsonc-parser.test.ts packages/utils/src/frontmatter.test.ts packages/hashline-core/src/hash-computation.test.ts packages/hashline-core/src/smoke-untested-modules.test.ts packages/rules-engine/src/index.test.ts packages/rules-engine/src/security-boundary.test.ts packages/agents-md-core/src/injector.test.ts packages/omo-codex/plugin/components/lsp/test/package-smoke.test.ts && node --test packages/omo-codex/plugin/test/*.test.mjs packages/omo-codex/scripts/install-cache-copy.test.mjs packages/omo-codex/scripts/install-cli-args.test.mjs packages/omo-codex/scripts/install-delegated-command.test.mjs packages/omo-codex/scripts/install-config-autonomous-features.test.mjs packages/omo-codex/scripts/install-config-autonomous.test.mjs packages/omo-codex/scripts/install-config-reasoning.test.mjs packages/omo-codex/scripts/install-config.test.mjs packages/omo-codex/scripts/install-hook-targets.test.mjs packages/omo-codex/scripts/install-project-local-cleanup.test.mjs packages/omo-codex/scripts/install-lazycodex-version-stamp.test.mjs packages/omo-codex/scripts/install-local-entrypoint.test.mjs packages/omo-codex/scripts/install-local-git-bash-preflight.test.mjs packages/omo-codex/scripts/install-local.test.mjs packages/omo-codex/scripts/install-marketplace-cache.test.mjs packages/omo-codex/scripts/install-mcp-context7-runtime.test.mjs packages/omo-codex/scripts/install-mcp-runtime.test.mjs packages/omo-codex/scripts/install-packaged-local.test.mjs packages/omo-codex/scripts/install-generated-bundle.test.mjs packages/omo-codex/scripts/install-agent-links.test.mjs packages/omo-codex/scripts/install-bin-links.test.mjs",
     "test:senpi": "node packages/omo-senpi/plugin/scripts/build-extension.mjs && node packages/omo-senpi/plugin/scripts/sync-skills.mjs && node packages/omo-senpi/plugin/scripts/embed-directive.mjs --check && bun test packages/omo-senpi",
     "test:windows-codex": "bun run test:codex",
     "build:git-bash-mcp": "bun run --cwd packages/git-bash-mcp build"
@@ -89823,18 +89866,18 @@ var package_default = {
     typescript: "^6.0.3"
   },
   optionalDependencies: {
-    "oh-my-opencode-darwin-arm64": "4.16.3",
-    "oh-my-opencode-darwin-x64": "4.16.3",
-    "oh-my-opencode-darwin-x64-baseline": "4.16.3",
-    "oh-my-opencode-linux-arm64": "4.16.3",
-    "oh-my-opencode-linux-arm64-musl": "4.16.3",
-    "oh-my-opencode-linux-x64": "4.16.3",
-    "oh-my-opencode-linux-x64-baseline": "4.16.3",
-    "oh-my-opencode-linux-x64-musl": "4.16.3",
-    "oh-my-opencode-linux-x64-musl-baseline": "4.16.3",
-    "oh-my-opencode-windows-arm64": "4.16.3",
-    "oh-my-opencode-windows-x64": "4.16.3",
-    "oh-my-opencode-windows-x64-baseline": "4.16.3"
+    "oh-my-opencode-darwin-arm64": "4.17.0",
+    "oh-my-opencode-darwin-x64": "4.17.0",
+    "oh-my-opencode-darwin-x64-baseline": "4.17.0",
+    "oh-my-opencode-linux-arm64": "4.17.0",
+    "oh-my-opencode-linux-arm64-musl": "4.17.0",
+    "oh-my-opencode-linux-x64": "4.17.0",
+    "oh-my-opencode-linux-x64-baseline": "4.17.0",
+    "oh-my-opencode-linux-x64-musl": "4.17.0",
+    "oh-my-opencode-linux-x64-musl-baseline": "4.17.0",
+    "oh-my-opencode-windows-arm64": "4.17.0",
+    "oh-my-opencode-windows-x64": "4.17.0",
+    "oh-my-opencode-windows-x64-baseline": "4.17.0"
   },
   overrides: {
     "@earendil-works/pi-agent-core": "0.80.3",
@@ -94916,10 +94959,10 @@ Trigger when ANY apply: user said "\uC5C4\uBC00" / "strictly" / "rigorously" / "
 
 Procedure (non-negotiable):
 1. Spawn a reviewer via \`task(category="ultrabrain", subagent_type="plan", load_skills=[...], run_in_background=false, prompt="<goal + scenarios + evidence + diff + notepad path>")\` \u2014 or any high-rigor reviewer agent available.
-2. Reviewer verdict is BINDING. There is no "false positive". Do not argue, minimise, or explain away.
-3. Fix every concern. Re-run the FULL scenario QA. Capture fresh evidence. Update notepad.
-4. Re-submit to the SAME reviewer. Loop until UNCONDITIONAL approval. "looks good but..." = REJECTION.
-5. Only on unconditional approval may you declare done.
+2. Verify each reviewer concern yourself. A concern blocks only when it names a success criterion the evidence fails; record concerns that cite no criterion as notes with a one-line reason \u2014 fixed or declined at your judgment.
+3. Fix every criterion-cited blocker. Re-run ONLY the scenario QA affected by the fix; capture fresh evidence for the delta. Update notepad.
+4. Re-submit to the SAME reviewer at most twice, passing only the delta diff, the blockers it cited, and the already-approved criteria marked out-of-scope. An approval whose only remaining items are notes counts as approval.
+5. On approval, declare done. If criterion-cited blockers remain after two re-reviews, stop and surface them to the user \u2014 do not loop further.
 
 ## ZERO TOLERANCE FAILURES
 - **NO Scope Reduction**: Never make "demo", "skeleton", "simplified", "basic" versions - deliver FULL implementation
@@ -95205,7 +95248,7 @@ If ANY answer is no \u2192 GO BACK AND DO IT. Do not claim completion.
 
 ### REVIEWER GATE (triggered, not optional)
 
-Trigger if user said "\uC5C4\uBC00"/"strictly"/"rigorously"/"properly review", or task touches 3+ files OR ran 20+ turns OR 30+ min, or refactor/migration/perf/security. Spawn a high-rigor reviewer via \`task\` with: goal, scenarios, evidence paths, full diff, notepad path. Verdict is BINDING. "looks good but..." = REJECTION. Fix every concern, re-run full scenario QA, capture fresh evidence, resubmit. Loop until UNCONDITIONAL approval.
+Trigger if user said "\uC5C4\uBC00"/"strictly"/"rigorously"/"properly review", or task touches 3+ files OR ran 20+ turns OR 30+ min, or refactor/migration/perf/security. Spawn a high-rigor reviewer via \`task\` with: goal, scenarios, evidence paths, full diff, notepad path. A concern blocks only when it names a success criterion the evidence fails; others are notes. Fix cited blockers, re-run the affected scenario QA, capture fresh delta evidence, and resubmit at most twice; an approval with only notes left counts as approval. Remaining cited blockers after two re-reviews go to the user.
 
 <MANUAL_QA_MANDATE>
 ### YOU MUST EXECUTE MANUAL QA. THIS IS NOT OPTIONAL. DO NOT SKIP THIS.
@@ -95455,7 +95498,7 @@ If QA starts a server, browser, tmux session, port, temp dir, or background proc
 
 Use a high-rigor reviewer when the task touches 3+ files, changes security/performance/migration behavior, lasts 30+ minutes, or the user asks for strict review.
 
-Reviewer verdict is binding. Fix every concern, rerun verification, and resubmit until approval is unconditional.
+A reviewer concern binds only when it cites a success criterion the evidence fails; other concerns are notes. Fix cited blockers, rerun the affected verification, and resubmit the delta at most twice; then surface remaining blockers to the user.
 
 ## ZERO TOLERANCE FAILURES
 
@@ -95649,7 +95692,7 @@ Name the exact tool + exact invocation per scenario (literal \`curl\` / \`send-k
 
 ## REVIEWER GATE (triggered)
 
-Trigger if user said "\uC5C4\uBC00"/"strictly"/"rigorously"/"properly review", or task touches 3+ files OR ran 20+ turns OR 30+ min, or it's a refactor/migration/perf/security change. Spawn a high-rigor reviewer via \`task\` with goal + scenarios + evidence + diff. Reviewer verdict is BINDING; "looks good but..." = rejection. Re-submit until UNCONDITIONAL approval before declaring done.
+Trigger if user said "\uC5C4\uBC00"/"strictly"/"rigorously"/"properly review", or task touches 3+ files OR ran 20+ turns OR 30+ min, or it's a refactor/migration/perf/security change. Spawn a high-rigor reviewer via \`task\` with goal + scenarios + evidence + diff. A concern blocks only when it cites a success criterion the evidence fails \u2014 others are notes. Fix cited blockers, re-run only the affected QA, and re-submit the delta at most twice; an approval with only notes left counts as approval. If cited blockers remain after two re-reviews, surface them to the user before declaring done.
 
 ## COMPLETION CRITERIA
 
@@ -148838,6 +148881,8 @@ call_omo_agent(subagent_type="librarian", prompt="I'm looking for proven impleme
 - MUST NOT: Create criteria requiring "user clicks/interacts..."
 - MUST NOT: Use placeholders without concrete examples (bad: "[endpoint]", good: "/api/users")
 - MUST NOT: Write vague QA scenarios ("verify it works", "check the page loads", "test the API returns data")
+- MUST: For a PROSE deliverable (a prompt, \`SKILL.md\`, rule, or markdown/instruction file), make QA a human/agent READ against the intended behavior, or assert only a machine-consumed value (a parsed field, a sentinel a runtime greps, a doc JSON sample through its real validator) \u2014 the file's wording has no behavioral seam
+- MUST NOT: Turn a prompt/doc change into a text-grep acceptance criterion (\`grep "<sentence>" SKILL.md\`, word/char counts, phrase presence/absence) \u2014 that pins a diff, not behavior, and blocks every legitimate edit
 
 ## Recommended Approach
 [1-2 sentence summary of how to proceed]
@@ -160695,6 +160740,18 @@ function parseRalphLoopArguments(rawArguments) {
 
 // packages/omo-opencode/src/plugin/command-execute-before.ts
 init_logger2();
+
+// packages/omo-opencode/src/plugin/stop-continuation.ts
+function stopContinuation(args) {
+  const { directory, hooks, sessionID } = args;
+  hooks.stopContinuationGuard?.stop?.(sessionID);
+  hooks.todoContinuationEnforcer?.cancelAllCountdowns();
+  hooks.ralphLoop?.cancelLoop(sessionID);
+  clearBoulderState(directory);
+  log2("[stop-continuation] All continuation mechanisms stopped", { sessionID });
+}
+
+// packages/omo-opencode/src/plugin/command-execute-before.ts
 var NATIVE_LOOP_TRIGGERED_FLAG = "__omoNativeLoopTriggered";
 function hasPartsOutput(value) {
   if (typeof value !== "object" || value === null)
@@ -160704,11 +160761,14 @@ function hasPartsOutput(value) {
   return Array.isArray(parts);
 }
 function createCommandExecuteBeforeHandler(args) {
-  const { hooks } = args;
+  const { directory, hooks } = args;
   return async (input, output) => {
     await hooks.autoSlashCommand?.["command.execute.before"]?.(input, output);
     const normalizedCommand = input.command.toLowerCase();
     const sessionID = input.sessionID;
+    if (normalizedCommand === "stop-continuation" && sessionID) {
+      stopContinuation({ directory, hooks, sessionID });
+    }
     if (hooks.ralphLoop && sessionID) {
       if (normalizedCommand === "ralph-loop" || normalizedCommand === "ulw-loop") {
         const parsedArguments = parseRalphLoopArguments(input.arguments || "");
@@ -160954,6 +161014,14 @@ function createChatMessageHandler3(args) {
     if (input.agent) {
       updateSessionAgent(input.sessionID, input.agent);
     }
+    const slashCommand = detectSlashCommand(extractPromptText3(output.parts));
+    if (slashCommand?.command === "stop-continuation") {
+      stopContinuation({
+        directory: ctx.directory,
+        hooks,
+        sessionID: input.sessionID
+      });
+    }
     const isFirstMessage = firstMessageVariantGate.shouldOverride(input.sessionID);
     if (isFirstMessage) {
       firstMessageVariantGate.markApplied(input.sessionID);
@@ -162941,13 +163009,7 @@ function createToolExecuteBeforeHandler3(args) {
       const command = rawName?.replace(/^\//, "").toLowerCase();
       const sessionID = input.sessionID || getMainSessionID();
       if (command === "stop-continuation" && sessionID) {
-        hooks.stopContinuationGuard?.stop(sessionID);
-        hooks.todoContinuationEnforcer?.cancelAllCountdowns();
-        hooks.ralphLoop?.cancelLoop(sessionID);
-        clearBoulderState(ctx.directory);
-        log2("[stop-continuation] All continuation mechanisms stopped", {
-          sessionID
-        });
+        stopContinuation({ directory: ctx.directory, hooks, sessionID });
       }
       const workStartingCommands = ["start-work", "ralph-loop", "ulw-loop"];
       if (workStartingCommands.includes(command ?? "") && sessionID) {
@@ -162981,6 +163043,7 @@ function createPluginInterface(args) {
     },
     "chat.headers": createChatHeadersHandler({ ctx }),
     "command.execute.before": createCommandExecuteBeforeHandler({
+      directory: ctx.directory,
       hooks
     }),
     "chat.message": createChatMessageHandler3({

package/dist/plugin/chat-message/types.d.ts

@@ -49,6 +49,9 @@ type RalphLoopHook = {
     resumeLoop?: (sessionID: string) => boolean;
     cancelLoop: (sessionID: string) => boolean | void;
 };
+type TodoContinuationEnforcerHook = {
+    cancelAllCountdowns: () => void;
+};
 export type ChatMessageHooks = {
     modelFallback?: ChatMessageHook | null;
     stopContinuationGuard?: StopContinuationGuard | null;
@@ -63,5 +66,6 @@ export type ChatMessageHooks = {
     hephaestusAgentsMdInjector?: ChatMessageHook | null;
     startWork?: ChatMessageHook | null;
     ralphLoop?: RalphLoopHook | null;
+    todoContinuationEnforcer?: TodoContinuationEnforcerHook | null;
 };
 export {};

package/dist/plugin/command-execute-before.d.ts

@@ -14,6 +14,7 @@ type CommandExecuteBeforeOutput = {
 };
 declare const NATIVE_LOOP_TRIGGERED_FLAG = "__omoNativeLoopTriggered";
 export declare function createCommandExecuteBeforeHandler(args: {
+    directory: string;
     hooks: CreatedHooks;
 }): (input: CommandExecuteBeforeInput, output: CommandExecuteBeforeOutput) => Promise<void>;
 export { NATIVE_LOOP_TRIGGERED_FLAG };

package/dist/plugin/stop-continuation.d.ts

@@ -0,0 +1,17 @@
+type StopContinuationHooks = {
+    readonly stopContinuationGuard?: {
+        readonly stop?: (sessionID: string) => void;
+    } | null;
+    readonly todoContinuationEnforcer?: {
+        readonly cancelAllCountdowns: () => void;
+    } | null;
+    readonly ralphLoop?: {
+        readonly cancelLoop: (sessionID: string) => boolean | void;
+    } | null;
+};
+export declare function stopContinuation(args: {
+    readonly directory: string;
+    readonly hooks: StopContinuationHooks;
+    readonly sessionID: string;
+}): void;
+export {};

package/dist/skills/remove-ai-slops/SKILL.md

@@ -62,7 +62,7 @@ The agent looks for these nine categories. The first three are stylistic, the ne
    **Hard rule**: only apply when behavior equivalence is obvious. Do NOT change algorithms with subtle correctness implications. Do NOT micro-optimize hot paths without a benchmark. If in doubt, SKIP.
 
 ### Behavior coverage
-9. **Missing tests** — behavior present in changed files that is not locked by any regression test. The fix is not to remove code but to ADD the narrowest test that pins the behavior.
+9. **Missing tests** — behavior present in changed files that is not locked by any regression test. The fix is not to remove code but to ADD the narrowest test that pins the behavior. EXCEPTION: a PROSE file (prompt, `SKILL.md`, rule, markdown) has no behavioral seam — do NOT add a text/word-count/phrase pin for it; that guards a diff, not behavior. Cover only a machine-consumed value (parsed field, sentinel a runtime greps, a doc JSON sample through its real validator) or leave it to review.
 
 ### Structural
 10. **Oversized modules** — any source file exceeding **250 pure LOC** (non-blank, non-comment lines). This is an architectural defect, not a style preference. Measure: `awk '!/^[[:space:]]*$/ && !/^[[:space:]]*(#|\/\/)/' <file> | wc -l`.
@@ -122,7 +122,7 @@ For each in-scope source file:
 
 1. Identify the public/observable behavior the file exposes (exported functions, HTTP handlers, CLI commands, classes used elsewhere).
 2. Check whether existing tests cover that behavior. Use `git grep` / project test conventions to find related test files.
-3. **If behavior is uncovered or weakly covered, write the narrowest regression test that pins current behavior BEFORE editing the file.** Tests should pin observable outputs, not implementation details.
+3. **If behavior is uncovered or weakly covered, write the narrowest regression test that pins current behavior BEFORE editing the file.** Tests should pin observable outputs, not implementation details. A PROSE file (prompt/`SKILL.md`/rule/markdown) is exempt — its wording is not behavior; skip the test and rely on review, or assert only a machine-consumed value.
 4. Run the test suite (or at minimum the relevant tests). They must be **green** before any cleanup begins.
 
 If you cannot establish a green baseline (e.g., test runner is broken), STOP and report. Do not proceed with cleanup on unverified ground.

package/dist/skills/start-work/SKILL.md

@@ -14,13 +14,16 @@ Translate any OpenCode-only tool name in an inherited example to its Codex equiv
 | OpenCode example | Codex tool to use |
 | --- | --- |
 | final-review `task(...)` | `multi_agent_v1.spawn_agent({"message":"TASK: act as a rigorous reviewer. ...","agent_type":"lazycodex-gate-reviewer","fork_context":false})` |
-| worker `task(...)` | `multi_agent_v1.spawn_agent({"message":"TASK: act as <role>. ...","fork_context":false})` |
+| worker `task(...)` | `multi_agent_v1.spawn_agent({"message":"TASK: act as <role>. ...","fork_context":false})` — for implementation workers add `agent_type: "lazycodex-worker-<low|medium|high>"` when the spawn schema exposes `agent_type` |
 | `background_output(task_id="...")` | `multi_agent_v1.wait_agent(...)` for mailbox signals |
 | `team_*(...)` | `multi_agent_v1.spawn_agent` + `multi_agent_v1.send_input` + `multi_agent_v1.wait_agent` + `multi_agent_v1.close_agent` |
 
 When translating `load_skills=[...]`, name the skills inside the spawned agent's `message`. If a code block below conflicts with this section, this section wins.
 
-Codex exposes ONE of two subagent tool surfaces per session; check your own tool list and route accordingly. If `multi_agent_v1.*` tools exist, use the table above as written. If instead a flat `spawn_agent` with a required `task_name` exists (`multi_agent_v2`), rewrite every `multi_agent_v1.*` example: `multi_agent_v1.spawn_agent({...,"fork_context":false})` becomes `spawn_agent({"task_name":"<lowercase_digits_underscores>","message":...,"agent_type":...,"fork_turns":"none"})` (`"all"` only when full parent history is truly required); `send_input` becomes `send_message`; do not call `close_agent`/`resume_agent` (finished agents end on their own; `followup_task` re-tasks one, `interrupt_agent` stops one); `wait_agent` takes only `timeout_ms` and returns on any child mailbox activity. `agent_type` works the same on both surfaces. If a code block below conflicts with this section, this section wins.
+Codex exposes ONE of two subagent tool surfaces per session; check your own tool list and route accordingly. If `multi_agent_v1.*` tools exist, use the table above as written. If instead a flat `spawn_agent` with a required `task_name` exists (`multi_agent_v2`), rewrite every `multi_agent_v1.*` example: `multi_agent_v1.spawn_agent({...,"fork_context":false})` becomes `spawn_agent({"task_name":"<lowercase_digits_underscores>","message":...,"agent_type":...,"fork_turns":"none"})` (`"all"` only when full parent history is truly required); `send_input` becomes `send_message`; do not call `close_agent`/`resume_agent` (finished agents end on their own; `followup_task` re-tasks one, `interrupt_agent` stops one); `wait_agent` takes only `timeout_ms` and returns on any child mailbox activity. On the v2 surface `agent_type` may be absent from the spawn schema — when absent, omit it and describe the role inside `message`. If a code block below conflicts with this section, this section wins.
+
+### Delegation by difficulty (Codex tier workers)
+When tier worker agents are installed (Codex), size each implementation lane by difficulty and pass the matching `agent_type` where the spawn schema exposes it: LOW (one-file fix, boilerplate, config/copy) -> `lazycodex-worker-low`; MEDIUM (standard feature, few files, known patterns) -> `lazycodex-worker-medium`; HIGH (new module, cross-module refactor, concurrency/security/migration) -> `lazycodex-worker-high`. Explorer/librarian research lanes keep their own roles. Difficulty (model power) is orthogonal to the LIGHT/HEAVY rigor tier in step 4 — judge each on its own facts. On spawn surfaces without `agent_type` (deployed v2), state the tier inside `message`.
 
 ## Codex Subagent Reliability
 

package/dist/skills/ulw-plan/scripts/scaffold-plan.mjs

@@ -221,7 +221,7 @@ Your next move: <fill - e.g. approve, or run a high-accuracy review>. Full execu
 ## Verification strategy
 > Zero human intervention - all verification is agent-executed.
 - Test decision: <TDD | tests-after | none> + framework
-- Evidence: .omo/evidence/task-<N>-${slug}.<ext>
+- Evidence: <attemptDir>/task-<N>-${slug}.<ext> (attemptDir = currentAttemptDir from 'omo ulw-loop status --json', .omo/evidence/ulw/<session>/<goalId>/a<attempt>; outside ulw-loop use .omo/evidence/)
 
 ## Execution strategy
 ### Parallel execution waves
@@ -239,7 +239,7 @@ Your next move: <fill - e.g. approve, or run a high-accuracy review>. Full execu
   Parallelization: Wave <N> | Blocked by: <...> | Blocks: <...>
   References (executor has NO interview context - be exhaustive): <src/path:lines>
   Acceptance criteria (agent-executable): <exact command or assertion>
-  QA scenarios (name the exact tool + invocation): happy + failure, Evidence .omo/evidence/task-1-${slug}.<ext>
+  QA scenarios (name the exact tool + invocation): happy + failure, Evidence <attemptDir>/task-1-${slug}.<ext>
   Commit: <Y/N> | <type>(<scope>): <summary>
 
 ## Final verification wave

package/dist/tui.js

@@ -20904,6 +20904,11 @@ var init_category_model_requirements = __esm(() => {
     },
     deep: {
       fallbackChain: [
+        {
+          providers: ["openai", "vercel"],
+          model: "gpt-5.6-terra",
+          variant: "xhigh"
+        },
         {
           providers: ["openai", "vercel"],
           model: "gpt-5.6-sol",
@@ -20967,6 +20972,11 @@ var init_category_model_requirements = __esm(() => {
     },
     "unspecified-low": {
       fallbackChain: [
+        {
+          providers: ["openai", "vercel"],
+          model: "gpt-5.6-luna",
+          variant: "xhigh"
+        },
         {
           providers: ["anthropic", "github-copilot", "opencode", "vercel"],
           model: "claude-sonnet-4-6"
@@ -21364,6 +21374,38 @@ var init_supplemental_entries = __esm(() => {
         output: 128000
       }
     },
+    "gpt-5.6-terra": {
+      id: "gpt-5.6-terra",
+      family: "gpt-mini",
+      reasoning: true,
+      temperature: false,
+      toolCall: true,
+      modalities: {
+        input: ["text", "image", "pdf"],
+        output: ["text"]
+      },
+      limit: {
+        context: 1050000,
+        input: 922000,
+        output: 128000
+      }
+    },
+    "gpt-5.6-luna": {
+      id: "gpt-5.6-luna",
+      family: "gpt-nano",
+      reasoning: true,
+      temperature: false,
+      toolCall: true,
+      modalities: {
+        input: ["text", "image", "pdf"],
+        output: ["text"]
+      },
+      limit: {
+        context: 1050000,
+        input: 922000,
+        output: 128000
+      }
+    },
     "gpt-5.5": {
       id: "gpt-5.5",
       family: "gpt",
@@ -21882,28 +21924,7 @@ var init_agent_display_names = __esm(() => {
   REVERSE_DISPLAY_NAMES = Object.fromEntries(Object.entries(AGENT_DISPLAY_NAMES).map(([key, displayName]) => [displayName.toLowerCase(), key]));
 });
 // packages/omo-opencode/src/tools/delegate-task/anthropic-categories.ts
-var UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND = `<Category_Context>
-You are working on tasks that don't fit specific categories but require moderate effort.
-
-<Selection_Gate>
-BEFORE selecting this category, VERIFY ALL conditions:
-1. Task does NOT fit: quick (trivial), visual-engineering (UI), ultrabrain (deep logic), artistry (creative), writing (docs)
-2. Task requires more than trivial effort but is NOT system-wide
-3. Scope is contained within a few files/modules
-
-If task fits ANY other category, DO NOT select unspecified-low.
-This is NOT a default choice - it's for genuinely unclassifiable moderate-effort work.
-</Selection_Gate>
-</Category_Context>
-
-<Caller_Warning>
-THIS CATEGORY USES A MID-TIER MODEL (claude-sonnet-4-6).
-
-**PROVIDE CLEAR STRUCTURE:**
-1. MUST DO: Enumerate required actions explicitly
-2. MUST NOT DO: State forbidden actions to prevent scope creep
-3. EXPECTED OUTPUT: Define concrete success criteria
-</Caller_Warning>`, UNSPECIFIED_HIGH_CATEGORY_PROMPT_APPEND = `<Category_Context>
+var UNSPECIFIED_HIGH_CATEGORY_PROMPT_APPEND = `<Category_Context>
 You are working on tasks that don't fit specific categories but require substantial effort.
 
 <Selection_Gate>
@@ -21919,12 +21940,6 @@ If task is unclassifiable but moderate-effort, use unspecified-low instead.
 </Category_Context>`, ANTHROPIC_CATEGORIES;
 var init_anthropic_categories = __esm(() => {
   ANTHROPIC_CATEGORIES = [
-    {
-      name: "unspecified-low",
-      config: { model: "anthropic/claude-sonnet-4-6" },
-      description: "Tasks that don't fit other categories, low effort required",
-      promptAppend: UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND
-    },
     {
       name: "unspecified-high",
       config: { model: "anthropic/claude-opus-4-7", variant: "max" },
@@ -22221,6 +22236,27 @@ EXPECTED OUTPUT:
 \`\`\`
 
 If your prompt lacks this structure, REWRITE IT before delegating.
+</Caller_Warning>`, UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND = `<Category_Context>
+You are working on tasks that don't fit specific categories but require moderate effort.
+
+<Selection_Gate>
+BEFORE selecting this category, VERIFY ALL conditions:
+1. Task does NOT fit: quick (trivial), visual-engineering (UI), ultrabrain (deep logic), artistry (creative), writing (docs)
+2. Task requires more than trivial effort but is NOT system-wide
+3. Scope is contained within a few files/modules
+
+If task fits ANY other category, DO NOT select unspecified-low.
+This is NOT a default choice - it's for genuinely unclassifiable moderate-effort work.
+</Selection_Gate>
+</Category_Context>
+
+<Caller_Warning>
+THIS CATEGORY USES A LIGHTWEIGHT MODEL (gpt-5.6-luna).
+
+**PROVIDE CLEAR STRUCTURE:**
+1. MUST DO: Enumerate required actions explicitly
+2. MUST NOT DO: State forbidden actions to prevent scope creep
+3. EXPECTED OUTPUT: Define concrete success criteria
 </Caller_Warning>`, OPENAI_CATEGORIES;
 var init_openai_categories = __esm(() => {
   init_types3();
@@ -22233,7 +22269,7 @@ var init_openai_categories = __esm(() => {
     },
     {
       name: "deep",
-      config: { model: "openai/gpt-5.6-sol", variant: "high" },
+      config: { model: "openai/gpt-5.6-terra", variant: "xhigh" },
       description: "Goal-oriented autonomous problem-solving on hairy problems requiring deep research. ONE goal + ONE deliverable per call \u2014 multiple goals must fan out as parallel `deep` calls, never bundled into one.",
       promptAppend: DEEP_CATEGORY_PROMPT_APPEND,
       resolvePromptAppend: resolveDeepCategoryPromptAppend
@@ -22243,6 +22279,12 @@ var init_openai_categories = __esm(() => {
       config: { model: "openai/gpt-5.4-mini" },
       description: "Trivial tasks - single file changes, typo fixes, simple modifications",
       promptAppend: QUICK_CATEGORY_PROMPT_APPEND
+    },
+    {
+      name: "unspecified-low",
+      config: { model: "openai/gpt-5.6-luna", variant: "xhigh" },
+      description: "Tasks that don't fit other categories, low effort required",
+      promptAppend: UNSPECIFIED_LOW_CATEGORY_PROMPT_APPEND
     }
   ];
 });