sneakoscope 0.7.42 → 0.7.44

package/README.md

@@ -42,7 +42,7 @@ sks selftest --mock
 
 | Area | What it does |
 | --- | --- |
-| CLI runtime | Bare `sks` opens or reuses the default tmux Codex CLI workspace. `sks tmux open` remains the explicit form for session/workspace flags, and `sks --mad` launches the high-reasoning auto-review profile. |
+| CLI runtime | Bare `sks` opens or reuses the default tmux Codex CLI workspace. `sks tmux open` remains the explicit form for session/workspace flags, and `sks --mad` launches the explicit full-access high-reasoning profile. |
 | Codex App commands | Installs generated skills so `$Team`, `$From-Chat-IMG`, `$DFix`, `$QA-LOOP`, `$PPT`, `$Goal`, `$DB`, `$Wiki`, `$Help`, and related routes are visible in prompt workflows. |
 | OpenClaw agents | Generates an OpenClaw skill package so OpenClaw agents can attach `sneakoscope-codex`, enable the `shell` tool, and discover/use SKS commands from the target repo root. |
 | Pipeline plans | Writes `pipeline-plan.json` for stateful routes so the runtime lane, kept stages, skipped stages, verification commands, and no-unrequested-fallback invariant are visible with `sks pipeline plan`. |
@@ -166,7 +166,7 @@ sks tmux check
 sks tmux status --once
 ```
 
-Bare `sks` creates or reuses the default named tmux session for Codex CLI and attaches to it in an interactive terminal. By default it launches Codex in the SKS fast-high runtime (`--model gpt-5.5 -c model_reasoning_effort="high"`) with a short animated SKS ASCII intro. Override with `SKS_CODEX_MODEL`, `SKS_CODEX_REASONING`, disable the default model profile with `SKS_CODEX_FAST_HIGH=0`, or disable the intro animation with `SKS_TMUX_LOGO_ANIMATION=0`. Use `sks tmux open` when you need explicit `--workspace` / `--session` flags, `sks tmux check` for readiness without launching, and `sks help` for CLI help. Use `--no-attach` or `SKS_TMUX_NO_AUTO_ATTACH=1` when you only want SKS to create/reuse the session and print the manual attach command.
+Bare `sks` creates or reuses the default named tmux session for Codex CLI and attaches to it in an interactive terminal. By default it launches Codex in the SKS fast-high runtime (`--model gpt-5.5 -c model_reasoning_effort="high"`) with a static SKS 3D ASCII intro inside tmux; the animated intro is reserved for non-tmux unauthenticated Codex launches and can be disabled with `SKS_TMUX_LOGO_ANIMATION=0`. Override the runtime with `SKS_CODEX_MODEL`, `SKS_CODEX_REASONING`, or disable the default model profile with `SKS_CODEX_FAST_HIGH=0`. Use `sks tmux open` when you need explicit `--workspace` / `--session` flags, `sks tmux check` for readiness without launching, and `sks help` for CLI help. Use `--no-attach` or `SKS_TMUX_NO_AUTO_ATTACH=1` when you only want SKS to create/reuse the session and print the manual attach command.
 
 Before opening tmux, SKS checks the installed Codex CLI against npm `@openai/codex@latest`. If a newer version exists, it asks `Y/n`; answering `y` updates automatically with `npm i -g @openai/codex@latest` and then opens tmux with the updated Codex CLI.
 
@@ -204,7 +204,7 @@ sks --mad
 sks --mad --yes
 ```
 
-This creates/uses the `sks-mad-high` Codex profile for a one-shot full-access, high-reasoning tmux session with `approval_policy = "on-request"` and `approvals_reviewer = "auto_review"`, then attaches to the session in an interactive terminal. It is scoped to that explicit command and does not change normal SKS/DB safety defaults. Repeat launches reuse the same named SKS MAD tmux session.
+This creates/uses the `sks-mad-high` Codex profile for a one-shot full-access, high-reasoning tmux session with `sandbox_mode = "danger-full-access"` and `approval_policy = "never"`, then launches Codex with `--sandbox danger-full-access --ask-for-approval never` and attaches to the session in an interactive terminal. It is scoped to that explicit command and does not change normal SKS/DB safety defaults. Repeat launches reuse the same named SKS MAD tmux session.
 
 MAD does not disable the pipeline contract: stages, executors, reviewers, and auto-review policy still must not invent unrequested fallback implementation code. If the requested path cannot be implemented, SKS should block with evidence rather than add substitute behavior.
 

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "sneakoscope",
   "displayName": "ㅅㅋㅅ",
-  "version": "0.7.42",
+  "version": "0.7.44",
   "description": "Sneakoscope Codex: database-safe Codex CLI/App harness with Team, Goal, AutoResearch, TriWiki, and Honest Mode.",
   "type": "module",
   "homepage": "https://github.com/mandarange/Sneakoscope-Codex#readme",

package/src/cli/main.mjs

@@ -1005,11 +1005,11 @@ async function madHighCommand(args = []) {
     return;
   }
   const profile = await enableMadHighProfile();
-  console.log(`SKS MAD auto-review profile ready: ${madHighProfileName()}`);
-  console.log('Scope: explicit tmux launch only; full access uses Codex auto_review approvals when approval prompts are raised.');
+  console.log(`SKS MAD full-access profile ready: ${madHighProfileName()}`);
+  console.log('Scope: explicit tmux launch only; Codex opens with danger-full-access sandbox and approval_policy=never.');
   const workspace = readOption(cleanArgs, '--workspace', readOption(cleanArgs, '--session', `sks-mad-${defaultTmuxSessionName(process.cwd())}`));
   return launchTmuxUi([...cleanArgs, '--workspace', workspace], {
-    codexArgs: ['--profile', profile.profile_name],
+    codexArgs: profile.launch_args,
     autoInstallTmux: !flag(args, '--no-auto-install-tmux'),
     conciseBlockers: true
   });
@@ -1927,6 +1927,7 @@ async function selftest() {
   if (postinstallConflictPrompt.code !== 0 || !String(postinstallConflictPrompt.stdout || '').includes('Goal: completely remove the conflicting Codex harnesses')) throw new Error('selftest failed: interactive postinstall prompt did not print cleanup prompt');
   const postinstallSetupTmp = tmpdir();
   await writeJsonAtomic(path.join(postinstallSetupTmp, 'package.json'), { name: 'postinstall-setup-smoke', version: '0.0.0' });
+  await ensureDir(path.join(postinstallSetupTmp, '.git'));
   const postinstallSetup = await runProcess(process.execPath, [path.join(packageRoot(), 'bin', 'sks.mjs'), 'postinstall'], { cwd: postinstallSetupTmp, env: { INIT_CWD: postinstallSetupTmp, HOME: path.join(postinstallSetupTmp, 'home'), SKS_SKIP_POSTINSTALL_SHIM: '1', SKS_SKIP_POSTINSTALL_CONTEXT7: '1', SKS_SKIP_POSTINSTALL_GETDESIGN: '1', SKS_SKIP_CLI_TOOLS: '1' }, timeoutMs: 30000, maxOutputBytes: 256 * 1024 });
   if (postinstallSetup.code !== 0) throw new Error(`selftest failed: postinstall setup exited ${postinstallSetup.code}: ${postinstallSetup.stderr}`);
   if (await exists(path.join(postinstallSetupTmp, '.agents', 'skills', 'agent-team', 'SKILL.md'))) throw new Error('selftest failed: postinstall installed deprecated agent-team fallback skill');
@@ -1964,6 +1965,7 @@ async function selftest() {
   if (!String(postinstallNoMarker.stdout || '').includes('no project marker found; auto-running global SKS runtime bootstrap')) throw new Error('selftest failed: no-marker postinstall did not report global runtime bootstrap');
   if (!(await exists(path.join(postinstallNoMarkerGlobalRoot, '.sneakoscope', 'manifest.json')))) throw new Error('selftest failed: no-marker postinstall did not bootstrap global runtime root');
   if (await exists(path.join(postinstallNoMarkerCwd, '.sneakoscope'))) throw new Error('selftest failed: no-marker postinstall polluted install cwd');
+  if (await exists(path.join(postinstallNoMarkerGlobalRoot, '.gitignore'))) throw new Error('selftest failed: global runtime bootstrap without project git wrote shared .gitignore');
   const bootstrapJsonTmp = tmpdir();
   await writeJsonAtomic(path.join(bootstrapJsonTmp, 'package.json'), { name: 'bootstrap-json-smoke', version: '0.0.0' });
   const bootstrapJson = await runProcess(process.execPath, [path.join(packageRoot(), 'bin', 'sks.mjs'), 'bootstrap', '--json'], { cwd: bootstrapJsonTmp, env: { HOME: path.join(bootstrapJsonTmp, 'home'), SKS_SKIP_POSTINSTALL_GLOBAL_SKILLS: '1', SKS_SKIP_CLI_TOOLS: '1' }, timeoutMs: 30000, maxOutputBytes: 256 * 1024 });
@@ -1987,9 +1989,9 @@ async function selftest() {
   const madProfilePath = path.join(tmp, 'mad-codex-config.toml');
   const madProfile = await enableMadHighProfile({ configPath: madProfilePath });
   const madProfileText = await safeReadText(madProfilePath);
-  if (madProfile.profile_name !== 'sks-mad-high' || !madProfileText.includes('sandbox_mode = "danger-full-access"') || !madProfileText.includes('approval_policy = "on-request"') || !madProfileText.includes('approvals_reviewer = "auto_review"') || !madProfileText.includes('model_reasoning_effort = "high"') || !madProfileText.includes('unrequested fallback implementation code')) throw new Error('selftest failed: MAD high profile is not full-access auto-review high with fallback-code guard');
+  if (madProfile.profile_name !== 'sks-mad-high' || !madProfileText.includes('sandbox_mode = "danger-full-access"') || !madProfileText.includes('approval_policy = "never"') || !madProfileText.includes('approvals_reviewer = "auto_review"') || !madProfile.launch_args.includes('--sandbox') || !madProfile.launch_args.includes('danger-full-access') || !madProfile.launch_args.includes('--ask-for-approval') || !madProfile.launch_args.includes('never') || !madProfileText.includes('model_reasoning_effort = "high"') || !madProfileText.includes('unrequested fallback implementation code')) throw new Error('selftest failed: MAD high profile is not Codex full-access high with fallback-code guard');
   if (!isMadHighLaunch(['--mad', '--high']) || isMadHighLaunch(['db', '--mad'])) throw new Error('selftest failed: MAD high launch flag parsing is not top-level only');
-  const workspacePlan = { session: 'sks-mad-selftest', root: tmp, codexArgs: ['--profile', 'sks-mad-high'] };
+  const workspacePlan = { session: 'sks-mad-selftest', root: tmp, codexArgs: madProfile.launch_args };
   const tmuxSyntax = runTmuxLaunchPlanSyntaxCheck(workspacePlan);
   if (!tmuxSyntax.ok || !tmuxSyntax.command.includes('tmux attach-session -t sks-mad-selftest')) throw new Error('selftest failed: MAD tmux attach plan is not stable by session name');
   const tmuxOpenArgs = buildTmuxOpenArgs(workspacePlan);
@@ -2405,7 +2407,8 @@ async function selftest() {
   if (!hookKoreanSksContext.includes('Route: $Team')) throw new Error('selftest failed: Korean implementation prompt did not promote to Team route');
   if (hookKoreanSksContext.includes('SKS answer-only pipeline active')) throw new Error('selftest failed: Korean implementation prompt still used answer-only pipeline');
   const hookKoreanSksState = await readJson(stateFile(hookKoreanSksTmp), {});
-  if (hookKoreanSksState.phase !== 'TEAM_CLARIFICATION_CONTRACT_SEALED' || hookKoreanSksState.implementation_allowed !== true || !hookKoreanSksState.ambiguity_gate_passed) throw new Error('selftest failed: Korean Team auto-seal');
+  if (hookKoreanSksState.phase !== 'TEAM_PARALLEL_ANALYSIS_SCOUTING' || hookKoreanSksState.implementation_allowed !== true || !hookKoreanSksState.ambiguity_gate_passed || !hookKoreanSksState.team_plan_ready) throw new Error('selftest failed: Korean Team auto-seal did not materialize Team');
+  if (!(await exists(path.join(missionDir(hookKoreanSksTmp, hookKoreanSksState.mission_id), 'team-plan.json')))) throw new Error('selftest failed: Korean Team auto-seal did not write team-plan.json');
   const hookPaymentTeamTmp = tmpdir();
   await initProject(hookPaymentTeamTmp, {});
   const hookPaymentTeamPayload = JSON.stringify({ cwd: hookPaymentTeamTmp, prompt: '$Team 결제 재시도 정책과 로그인 세션 만료 버그 수정 executor:2 reviewer:1 user:1' });
@@ -2416,9 +2419,10 @@ async function selftest() {
   if (!hookPaymentTeamContext.includes('Ambiguity gate auto-sealed')) throw new Error('selftest failed: predictable payment/auth Team prompt did not auto-seal');
   if (hookPaymentTeamContext.includes('PAYMENT_RETRY_POLICY') || hookPaymentTeamContext.includes('AUTH_PROTOCOL_CHANGE_ALLOWED')) throw new Error('selftest failed: predictable payment/auth policy defaults were asked instead of inferred');
   const hookPaymentTeamState = await readJson(stateFile(hookPaymentTeamTmp), {});
-  if (hookPaymentTeamState.phase !== 'TEAM_CLARIFICATION_CONTRACT_SEALED' || hookPaymentTeamState.implementation_allowed !== true || !hookPaymentTeamState.ambiguity_gate_passed) throw new Error('selftest failed: predictable payment/auth Team state was not executable after auto-seal');
+  if (hookPaymentTeamState.phase !== 'TEAM_PARALLEL_ANALYSIS_SCOUTING' || hookPaymentTeamState.implementation_allowed !== true || !hookPaymentTeamState.ambiguity_gate_passed || !hookPaymentTeamState.team_plan_ready) throw new Error('selftest failed: predictable payment/auth Team did not materialize after auto-seal');
   const hookPaymentTeamSchema = await readJson(path.join(missionDir(hookPaymentTeamTmp, hookPaymentTeamState.mission_id), 'required-answers.schema.json'));
   if (hookPaymentTeamSchema.slots.length !== 0 || hookPaymentTeamSchema.inferred_answers?.PAYMENT_RETRY_POLICY === undefined || hookPaymentTeamSchema.inferred_answers?.AUTH_SESSION_EXPIRED_BEHAVIOR === undefined) throw new Error('selftest failed: predictable payment/auth defaults were not recorded as inferred answers');
+  if (!(await exists(path.join(missionDir(hookPaymentTeamTmp, hookPaymentTeamState.mission_id), 'team-plan.json')))) throw new Error('selftest failed: predictable payment/auth Team auto-seal did not write team-plan.json');
   const hookTeamTmp = tmpdir();
   await initProject(hookTeamTmp, {});
   const hookTeamPayload = JSON.stringify({ cwd: hookTeamTmp, prompt: '$Team 발표자료 만들어줘 executor:2 reviewer:1 user:1' });
@@ -2626,6 +2630,7 @@ async function selftest() {
   if (!codexConfigText.includes('multi_agent = true')) throw new Error('selftest failed: multi_agent not enabled');
   if (!hasContext7ConfigText(codexConfigText)) throw new Error('selftest failed: Context7 MCP not configured');
   if (!codexConfigText.includes('[profiles.sks-task-low]') || !codexConfigText.includes('[profiles.sks-task-medium]') || !codexConfigText.includes('[profiles.sks-logic-high]') || !codexConfigText.includes('[profiles.sks-fast-high]') || !codexConfigText.includes('[profiles.sks-research-xhigh]') || !codexConfigText.includes('[profiles.sks-mad-high]')) throw new Error('selftest failed: GPT-5.5 reasoning profiles not configured');
+  if (!/\[profiles\.sks-mad-high\][\s\S]*?approval_policy = "never"[\s\S]*?sandbox_mode = "danger-full-access"/.test(codexConfigText)) throw new Error('selftest failed: generated sks-mad-high profile is not full access');
   if (!codexConfigText.includes('[agents.analysis_scout]')) throw new Error('selftest failed: analysis_scout agent not configured');
   if (!codexConfigText.includes('[agents.team_consensus]')) throw new Error('selftest failed: team_consensus agent not configured');
   const preservedConfigTmp = tmpdir();
@@ -2633,7 +2638,7 @@ async function selftest() {
   await writeTextAtomic(path.join(preservedConfigTmp, '.codex', 'config.toml'), '[features]\nfast_mode_ui = true\n\n[user.fast_mode]\nvisible = true\n');
   await initProject(preservedConfigTmp, {});
   const preservedConfig = await safeReadText(path.join(preservedConfigTmp, '.codex', 'config.toml'));
-  if (!preservedConfig.includes('fast_mode_ui = true') || !preservedConfig.includes('[user.fast_mode]') || !preservedConfig.includes('visible = true')) throw new Error('selftest failed: Codex config merge dropped user Fast mode settings');
+  if (!preservedConfig.includes('fast_mode_ui = true') || !preservedConfig.includes('[user.fast_mode]') || !preservedConfig.includes('visible = true') || !preservedConfig.includes('enabled = true') || !preservedConfig.includes('default_profile = "sks-fast-high"')) throw new Error('selftest failed: Codex config merge dropped or failed to enable Fast mode settings');
   if (!preservedConfig.includes('codex_hooks = true') || !preservedConfig.includes('[profiles.sks-fast-high]')) throw new Error('selftest failed: Codex config merge did not add SKS managed settings');
   const autoReviewHome = path.join(tmp, 'auto-review-home');
   const autoReviewEnv = { HOME: autoReviewHome };
@@ -2988,6 +2993,10 @@ async function selftest() {
   if (buttonUxSlotIds.length) throw new Error(`selftest failed: clear small UI work should auto-seal, got ${buttonUxSlotIds.join(',')}`);
   if (buttonUxSchema.inferred_answers.UI_STATE_BEHAVIOR !== 'infer_from_task_context_and_existing_design_system; preserve existing loading/error/empty/retry behavior unless explicitly requested; add only standard states required by the touched surface') throw new Error('selftest failed: UI state default inference missing');
   if (buttonUxSchema.inferred_answers.VISUAL_REGRESSION_REQUIRED !== 'yes_if_available') throw new Error('selftest failed: visual regression default inference missing');
+  const predictableAuthCliSchema = buildQuestionSchema('회전 아스키 아트는 제일 처음 인증 안됐을때만 codex cli처럼 애니메이션으로 보이게 하고 tmux에서는 정적 3d 아스키 아트로 보여줘');
+  const predictableAuthCliSlotIds = predictableAuthCliSchema.slots.map((s) => s.id);
+  if (predictableAuthCliSlotIds.length) throw new Error(`selftest failed: clear auth-worded CLI rendering work should auto-seal, got ${predictableAuthCliSlotIds.join(',')}`);
+  if (!predictableAuthCliSchema.inferred_answers.RISK_BOUNDARY?.includes('no destructive commands or live data writes')) throw new Error('selftest failed: predictable auth-worded CLI work did not infer conservative risk boundary');
   const vagueSchema = buildQuestionSchema('뭔가 개선해줘');
   const vagueSlotIds = vagueSchema.slots.map((s) => s.id);
   if (!vagueSlotIds.includes('INTENT_TARGET') || vagueSlotIds.includes('GOAL_PRECISE') || vagueSlotIds.includes('ACCEPTANCE_CRITERIA')) throw new Error(`selftest failed: vague work should ask dynamic intent questions only, got ${vagueSlotIds.join(',')}`);
@@ -3034,6 +3043,7 @@ async function selftest() {
   if (!pptHtml.includes('<html') || pptHtml.includes('gradient')) throw new Error('selftest failed: PPT HTML artifact missing or over-designed');
   const pptStyleTokens = await readJson(path.join(pptMission.dir, 'ppt-style-tokens.json'));
   if (pptStyleTokens.design_policy?.design_ssot?.authority !== DESIGN_SYSTEM_SSOT.authority_file || !pptStyleTokens.design_policy?.source_inputs?.some((entry) => entry.url === AWESOME_DESIGN_MD_REFERENCE.url && entry.role === 'source_input_for_ssot') || !pptStyleTokens.design_policy?.anti_generic_ai_style) throw new Error('selftest failed: PPT style tokens missing fused design SSOT/source-input anti-generic policy');
+  if (!pptStyleTokens.design_policy?.design_reference_selection?.primary?.id?.startsWith('awesome-design-md:') || !pptStyleTokens.design_policy?.design_reference_selection?.selected_sources?.length || !pptStyleTokens.layout?.composition || !pptStyleTokens.layout?.treatment) throw new Error('selftest failed: PPT style tokens did not select and apply a concrete awesome-design-md reference profile');
   if (JSON.stringify(pptStyleTokens.design_policy?.pipeline_allowlist?.required_skills || []) !== JSON.stringify(PPT_PIPELINE_SKILL_ALLOWLIST) || !pptStyleTokens.design_policy?.pipeline_allowlist?.ignore_installed_out_of_pipeline_skills || !(pptStyleTokens.design_policy?.pipeline_allowlist?.ignored_design_skills_even_if_installed || []).includes('design-artifact-expert') || !/AI-like/.test(pptStyleTokens.design_policy?.pipeline_allowlist?.anti_ai_design_goal || '')) throw new Error('selftest failed: PPT style tokens missing skill/MCP allowlist enforcement');
   const audienceScript = pptHtml.match(/id="ppt-audience-strategy">([^<]+)<\/script>/);
   if (!audienceScript) throw new Error('selftest failed: PPT HTML missing audience strategy script data');

package/src/core/auto-review.mjs

@@ -64,7 +64,7 @@ export async function enableMadHighProfile(opts = {}) {
   let next = upsertTable(current, `profiles.${MAD_HIGH_PROFILE}`, [
     `[profiles.${MAD_HIGH_PROFILE}]`,
     'model = "gpt-5.5"',
-    'approval_policy = "on-request"',
+    'approval_policy = "never"',
     `approvals_reviewer = "${AUTO_REVIEW_REVIEWER}"`,
     'sandbox_mode = "danger-full-access"',
     'model_reasoning_effort = "high"'
@@ -75,9 +75,9 @@ export async function enableMadHighProfile(opts = {}) {
   return {
     config_path: configPath,
     profile_name: MAD_HIGH_PROFILE,
-    launch_args: ['--profile', MAD_HIGH_PROFILE],
+    launch_args: ['--profile', MAD_HIGH_PROFILE, '--sandbox', 'danger-full-access', '--ask-for-approval', 'never'],
     sandbox_mode: 'danger-full-access',
-    approval_policy: 'on-request',
+    approval_policy: 'never',
     approvals_reviewer: AUTO_REVIEW_REVIEWER,
     model_reasoning_effort: 'high',
     scope: 'explicit_launch_only'

package/src/core/fsx.mjs

@@ -5,7 +5,7 @@ import os from 'node:os';
 import crypto from 'node:crypto';
 import { spawn } from 'node:child_process';
 
-export const PACKAGE_VERSION = '0.7.42';
+export const PACKAGE_VERSION = '0.7.44';
 export const DEFAULT_PROCESS_TAIL_BYTES = 256 * 1024;
 export const DEFAULT_PROCESS_TIMEOUT_MS = 30 * 60 * 1000;
 

package/src/core/init.mjs

@@ -113,8 +113,9 @@ export async function initProject(root, opts = {}) {
     '.sneakoscope/state', '.sneakoscope/missions', '.sneakoscope/db', '.sneakoscope/bus', '.sneakoscope/hproof', '.sneakoscope/db', '.sneakoscope/wiki', '.sneakoscope/skills', '.sneakoscope/memory/q0_raw', '.sneakoscope/memory/q1_evidence', '.sneakoscope/memory/q2_facts', '.sneakoscope/memory/q3_tags', '.sneakoscope/memory/q4_bits', '.sneakoscope/gx/cartridges', '.sneakoscope/model/fingerprints', '.sneakoscope/genome/candidates', '.sneakoscope/trajectories/raw', '.sneakoscope/locks', '.sneakoscope/tmp', '.sneakoscope/arenas', '.sneakoscope/reports', '.codex', '.codex/agents', '.agents/skills'
   ];
   for (const d of dirs) await ensureDir(path.join(root, d));
+  const sharedIgnoreWanted = !localOnly && await shouldWriteSharedGitIgnore(root, installScope);
   const localExclude = localOnly ? await ensureLocalOnlyGitExclude(root) : null;
-  const sharedIgnore = localOnly ? null : await ensureSharedGitIgnore(root);
+  const sharedIgnore = sharedIgnoreWanted ? await ensureSharedGitIgnore(root) : null;
   if (localExclude?.path) created.push(`${path.relative(root, localExclude.path)} local-only excludes`);
   if (sharedIgnore?.changed) created.push(`${path.relative(root, sharedIgnore.path)} SKS generated files ignore`);
 
@@ -429,6 +430,10 @@ function mergeManagedCodexConfigToml(existingContent = '') {
   let next = String(existingContent || '').trimEnd();
   next = upsertTomlTableKey(next, 'features', 'codex_hooks = true');
   next = upsertTomlTableKey(next, 'features', 'multi_agent = true');
+  next = upsertTomlTableKey(next, 'features', 'fast_mode_ui = true');
+  next = upsertTomlTableKey(next, 'user.fast_mode', 'visible = true');
+  next = upsertTomlTableKey(next, 'user.fast_mode', 'enabled = true');
+  next = upsertTomlTableKey(next, 'user.fast_mode', 'default_profile = "sks-fast-high"');
   next = upsertTomlTableKey(next, 'agents', 'max_threads = 6');
   next = upsertTomlTableKey(next, 'agents', 'max_depth = 1');
   for (const block of managedCodexConfigBlocks()) {
@@ -452,7 +457,7 @@ function managedCodexConfigBlocks() {
     { table: 'profiles.sks-research-xhigh', text: profileConfigBlock('sks-research-xhigh', 'xhigh') },
     { table: 'profiles.sks-research', text: profileConfigBlock('sks-research', 'xhigh', { approval: 'never' }) },
     { table: 'profiles.sks-team', text: profileConfigBlock('sks-team', 'high') },
-    { table: 'profiles.sks-mad-high', text: profileConfigBlock('sks-mad-high', 'high', { sandbox: 'danger-full-access', approvalsReviewer: 'auto_review' }) },
+    { table: 'profiles.sks-mad-high', text: profileConfigBlock('sks-mad-high', 'high', { approval: 'never', sandbox: 'danger-full-access', approvalsReviewer: 'auto_review' }) },
     {
       table: 'auto_review',
       text: '[auto_review]\npolicy = "Deny destructive database operations, credential exfiltration, persistent security weakening, broad file deletion, writes outside the workspace, and unrequested fallback implementation code unless explicitly authorized by the user or sealed decision contract."'
@@ -613,6 +618,13 @@ async function ensureSharedGitIgnore(root) {
   return { path: ignorePath, patterns, changed: true };
 }
 
+async function shouldWriteSharedGitIgnore(root, installScope) {
+  if (normalizeInstallScope(installScope) === 'project') return true;
+  if (await exists(path.join(root, '.git'))) return true;
+  if (await exists(path.join(root, '.gitignore'))) return true;
+  return false;
+}
+
 async function ensureLocalOnlyGitExclude(root) {
   const gitDir = await resolveGitDir(root);
   if (!gitDir) return { path: null, patterns: [] };

package/src/core/pipeline.mjs

@@ -504,10 +504,14 @@ async function prepareClarificationGate(root, route, task, required, opts = {})
   if (schema.slots.length === 0) {
     await writeJsonAtomic(path.join(dir, 'answers.json'), schema.inferred_answers || {});
     const result = await sealContract(dir, mission);
-    const plan = await writePipelinePlan(dir, { missionId: id, route, task, required, ambiguity: { required: true, slots: 0, auto_sealed: result.ok, passed: result.ok, contract_hash: result.contract?.sealed_hash || null } });
+    const materialized = result.ok && route?.id === 'Team'
+      ? await materializeAutoSealedTeam(root, id, dir, route, task, result.contract?.sealed_hash || null)
+      : {};
+    const effectiveTask = materialized.prompt || task;
+    const plan = await writePipelinePlan(dir, { missionId: id, route, task: effectiveTask, required, ambiguity: { required: true, slots: 0, auto_sealed: result.ok, passed: result.ok, contract_hash: result.contract?.sealed_hash || null } });
     await appendJsonl(path.join(dir, 'events.jsonl'), { ts: nowIso(), type: 'route.clarification.auto_sealed', route: route.id, slots: 0, ok: result.ok });
-    await setCurrent(root, routeState(id, route, result.ok ? `${route.mode}_CLARIFICATION_CONTRACT_SEALED` : `${route.mode}_CLARIFICATION_AWAITING_ANSWERS`, required, {
-      prompt: task,
+    await setCurrent(root, routeState(id, route, result.ok ? (materialized.phase || `${route.mode}_CLARIFICATION_CONTRACT_SEALED`) : `${route.mode}_CLARIFICATION_AWAITING_ANSWERS`, required, {
+      prompt: effectiveTask,
       questions_allowed: false,
       implementation_allowed: result.ok,
       clarification_required: false,
@@ -517,8 +521,10 @@ async function prepareClarificationGate(root, route, task, required, opts = {})
       pipeline_plan_ready: validatePipelinePlan(plan).ok,
       pipeline_plan_path: PIPELINE_PLAN_ARTIFACT,
       original_stop_gate: route.stopGate,
-      stop_gate: route.stopGate
+      stop_gate: route.stopGate,
+      ...(materialized.state || {})
     }));
+    const materializedLine = materialized.phase ? `\nTeam runtime artifacts were materialized immediately; state advanced to ${materialized.phase}.` : '';
     return {
       route,
       additionalContext: `${promptPipelineContext(task, route)}
@@ -528,6 +534,7 @@ Mission: ${id}
 Decision contract: .sneakoscope/missions/${id}/decision-contract.json
 Resolved answers: .sneakoscope/missions/${id}/resolved-answers.json
 Pipeline plan: .sneakoscope/missions/${id}/${PIPELINE_PLAN_ARTIFACT}
+${materializedLine}
 Next atomic action: continue the original route lifecycle with the sealed decision-contract.json.`
     };
   }
@@ -581,6 +588,91 @@ function applyMadSksAuthorizationToSchema(schema = {}) {
   return schema;
 }
 
+async function materializeAutoSealedTeam(root, id, dir, route, task, contractHash = null) {
+  const spec = parseTeamSpecText(task);
+  const cleanTask = spec.prompt || task;
+  const fromChatImgRequired = hasFromChatImgSignal(cleanTask);
+  const { agentSessions, roleCounts, roster } = spec;
+  const plan = {
+    schema_version: 1,
+    mission_id: id,
+    task: cleanTask,
+    agent_session_count: agentSessions,
+    default_agent_session_count: 3,
+    role_counts: roleCounts,
+    session_policy: `Use at most ${agentSessions} subagent sessions at a time; the parent orchestrator is not counted.`,
+    bundle_size: roster.bundle_size,
+    roster,
+    contract_hash: contractHash,
+    team_model: {
+      phases: ['parallel_analysis_scouts', 'triwiki_stage_refresh', 'debate_team', 'runtime_task_graph', 'development_team', 'review'],
+      analysis_team: `Read-only parallel scouting with exactly ${roster.bundle_size} analysis_scout_N agents.`,
+      debate_team: `Read-only role debate with exactly ${roster.bundle_size} participants.`,
+      development_team: `Fresh parallel development bundle with exactly ${roster.bundle_size} executor_N developers implementing disjoint slices.`
+    },
+    context_tracking: triwikiContextTracking(),
+    team_runtime: teamRuntimePlanMetadata(),
+    phases: [
+      { id: 'team_roster_confirmation', goal: `Materialize the Team roster from default SKS counts or explicit user counts, write team-roster.json, and surface role counts ${formatRoleCounts(roleCounts)}.`, agents: ['parent_orchestrator'], output: 'team-roster.json' },
+      { id: 'parallel_analysis_scouting', goal: `Read TriWiki context, then spawn exactly ${roster.bundle_size} read-only analysis_scout_N agents in parallel. ${fromChatImgRequired ? `From-Chat-IMG active: ${CODEX_COMPUTER_USE_ONLY_POLICY}` : 'From-Chat-IMG inactive: do not assume ordinary images are chat captures.'}`, agents: roster.analysis_team.map((agent) => agent.id), max_parallel_subagents: agentSessions, write_policy: 'read-only' },
+      { id: 'triwiki_refresh', goal: `Refresh or pack TriWiki and run ${triwikiContextTracking().validate_command}.`, agents: ['parent_orchestrator'], output: '.sneakoscope/wiki/context-pack.json' },
+      { id: 'planning_debate', goal: 'Run read-only planning debate, map constraints and implementation slices, then seal one objective.', agents: roster.debate_team.map((agent) => agent.id) },
+      { id: 'runtime_task_graph_compile', goal: `Compile the agreed Team plan into ${TEAM_GRAPH_ARTIFACT}, ${TEAM_RUNTIME_TASKS_ARTIFACT}, ${TEAM_DECOMPOSITION_ARTIFACT}, and ${TEAM_INBOX_DIR}.`, agents: ['parent_orchestrator'], output: [TEAM_GRAPH_ARTIFACT, TEAM_RUNTIME_TASKS_ARTIFACT, TEAM_DECOMPOSITION_ARTIFACT, TEAM_INBOX_DIR] },
+      { id: 'parallel_implementation', goal: `Close debate agents, then spawn a fresh ${roster.bundle_size}-person executor development team with non-overlapping write ownership.`, agents: roster.development_team.map((agent) => agent.id) },
+      { id: 'review_integration', goal: 'Review, integrate, verify, and record evidence before final.', agents: roster.validation_team.map((agent) => agent.id) },
+      { id: 'session_cleanup', goal: `Close or account for Team subagent sessions and write ${TEAM_SESSION_CLEANUP_ARTIFACT}.`, agents: ['parent_orchestrator'] }
+    ],
+    live_visibility: {
+      markdown: 'team-live.md',
+      transcript: 'team-transcript.jsonl',
+      dashboard: 'team-dashboard.json',
+      tmux: 'CLI Team entrypoints open tmux live lanes for the visible Team agent budget when tmux is available.',
+      commands: ['sks team status latest', 'sks team log latest', 'sks team tail latest', 'sks team watch latest', 'sks team lane latest --agent <name> --follow']
+    },
+    required_artifacts: ['team-roster.json', 'team-analysis.md', ...(fromChatImgRequired ? [FROM_CHAT_IMG_COVERAGE_ARTIFACT, FROM_CHAT_IMG_CHECKLIST_ARTIFACT, FROM_CHAT_IMG_TEMP_TRIWIKI_ARTIFACT, FROM_CHAT_IMG_QA_LOOP_ARTIFACT] : []), 'team-consensus.md', ...teamRuntimeRequiredArtifacts(), 'team-review.md', 'team-gate.json', TEAM_SESSION_CLEANUP_ARTIFACT, 'reflection.md', 'reflection-gate.json', 'team-live.md', 'team-transcript.jsonl', 'team-dashboard.json', '.sneakoscope/wiki/context-pack.json', 'context7-evidence.jsonl']
+  };
+  await writeJsonAtomic(path.join(dir, 'team-plan.json'), plan);
+  await writeJsonAtomic(path.join(dir, 'team-roster.json'), { schema_version: 1, mission_id: id, role_counts: roleCounts, agent_sessions: agentSessions, bundle_size: roster.bundle_size, roster, confirmed: true, source: 'auto_sealed_team_spec' });
+  const contextTracking = triwikiContextTracking();
+  await writeTextAtomic(path.join(dir, 'team-workflow.md'), `# SKS Team Workflow\n\nTask: ${cleanTask}\n\nAgent session budget: ${agentSessions}\nBundle size: ${roster.bundle_size}\nRole counts: ${formatRoleCounts(roleCounts)}\nContext tracking: ${contextTracking.ssot} SSOT, ${contextTracking.default_pack}.\n\nAuto-sealed ambiguity gate: no user question was required. Continue directly with Team scouting, debate, runtime task graph, implementation, review, cleanup, reflection, and Honest Mode.\n`);
+  await initTeamLive(id, dir, cleanTask, { agentSessions, roleCounts, roster });
+  const runtime = await writeTeamRuntimeArtifacts(dir, plan, { contractHash });
+  await writeMemorySweepReport(root, dir, { missionId: id }).catch(() => null);
+  await writeSkillForgeReport(dir, { mission_id: id, route: 'team', task_signature: cleanTask }).catch(() => null);
+  await writeMistakeMemoryReport(dir, { mission_id: id, route: 'team', task: cleanTask }).catch(() => null);
+  await writeCodeStructureReport(root, dir, { missionId: id, exception: 'Team auto-seal records split-review risk; extraction happens only when the mission scope includes the touched file.' }).catch(() => null);
+  await writeJsonAtomic(path.join(dir, 'team-gate.json'), {
+    passed: false,
+    team_roster_confirmed: true,
+    analysis_artifact: false,
+    triwiki_refreshed: false,
+    triwiki_validated: false,
+    consensus_artifact: false,
+    ...runtime.gate_fields,
+    implementation_team_fresh: false,
+    review_artifact: false,
+    integration_evidence: false,
+    session_cleanup: false,
+    context7_evidence: false,
+    ...(fromChatImgRequired ? { from_chat_img_required: true, from_chat_img_request_coverage: false } : {}),
+    contract_hash: contractHash
+  });
+  await appendJsonl(path.join(dir, 'events.jsonl'), { ts: nowIso(), type: 'team.materialized_after_auto_sealed_ambiguity_gate', route: route.id, bundle_size: roster.bundle_size, agent_sessions: agentSessions });
+  return {
+    phase: 'TEAM_PARALLEL_ANALYSIS_SCOUTING',
+    prompt: cleanTask,
+    state: {
+      agent_sessions: agentSessions,
+      role_counts: roleCounts,
+      team_roster_confirmed: true,
+      team_plan_ready: true,
+      team_graph_ready: runtime.ok,
+      team_live_ready: true,
+      from_chat_img_required: fromChatImgRequired
+    }
+  };
+}
+
 async function prepareTeam(root, route, task, required) {
   const spec = parseTeamSpecText(task);
   const cleanTask = spec.prompt || task;

package/src/core/ppt.mjs

@@ -16,6 +16,99 @@ export const PPT_CLEANUP_REPORT_ARTIFACT = 'ppt-cleanup-report.json';
 export const PPT_PARALLEL_REPORT_ARTIFACT = 'ppt-parallel-report.json';
 export const PPT_TEMP_DIR = 'ppt-tmp';
 
+const PPT_DESIGN_REFERENCE_PROFILES = Object.freeze([
+  {
+    id: 'awesome-design-md:ibm',
+    name: 'IBM Carbon enterprise',
+    source_url: 'https://raw.githubusercontent.com/VoltAgent/awesome-design-md/main/design-md/ibm/DESIGN.md',
+    source_summary: 'enterprise Carbon-style system: white surfaces, charcoal text, IBM Blue as the single accent, flat square tiles, thin rules, no shadow',
+    keywords: ['enterprise', 'b2b', 'investor', 'vc', 'strategy', 'proposal', 'board', 'finance', 'risk', 'compliance', '운영', '투자', '의사결정', '리스크', '전략'],
+    tokens: {
+      bg: '#ffffff',
+      text: '#161616',
+      muted: '#525252',
+      primary: '#0f62fe',
+      accent: '#393939',
+      surface: '#f4f4f4',
+      rule: '#e0e0e0',
+      display_px: 64,
+      body_px: 28,
+      caption_px: 15,
+      line_height: 1.36,
+      radius_px: 2,
+      treatment: 'flat_thin_rules_no_shadow',
+      composition: 'enterprise_evidence_grid',
+      mono_label: 'uppercase technical labels, sparse blue accent, source-visible rows'
+    },
+    applied_rules: [
+      'use white/charcoal enterprise canvas',
+      'reserve IBM Blue for one decision/action accent',
+      'prefer thin rules and square evidence rows over decorative cards',
+      'avoid shadows, gradients, and ornamental surfaces'
+    ]
+  },
+  {
+    id: 'awesome-design-md:vercel',
+    name: 'Vercel developer infrastructure',
+    source_url: 'https://raw.githubusercontent.com/VoltAgent/awesome-design-md/main/design-md/vercel/DESIGN.md',
+    source_summary: 'developer-infrastructure minimalism: white canvas, near-black type, shadow-as-border, mono technical labels, functional blue/red/pink workflow accents',
+    keywords: ['developer', 'devtools', 'api', 'sdk', 'cloud', 'infra', 'saas', 'technical', 'codex', 'ai', 'agent', '배포', '개발자', '기술', '자동화'],
+    tokens: {
+      bg: '#ffffff',
+      text: '#171717',
+      muted: '#4d4d4d',
+      primary: '#0072f5',
+      accent: '#ff5b4f',
+      surface: '#fafafa',
+      rule: '#ebebeb',
+      display_px: 66,
+      body_px: 28,
+      caption_px: 14,
+      line_height: 1.34,
+      radius_px: 8,
+      treatment: 'shadow_as_border_minimal_depth',
+      composition: 'technical_pipeline_grid',
+      mono_label: 'mono labels, workflow accent only when it clarifies sequence'
+    },
+    applied_rules: [
+      'use near-black text on a white technical canvas',
+      'show structure through shadow-as-border or one-pixel rules',
+      'use mono labels for sources and technical evidence',
+      'keep color functional rather than decorative'
+    ]
+  },
+  {
+    id: 'awesome-design-md:linear',
+    name: 'Linear precision operations',
+    source_url: 'https://github.com/VoltAgent/awesome-design-md',
+    source_summary: 'ultra-minimal precise product-management system: restrained neutral surfaces, exact spacing, one controlled purple accent',
+    keywords: ['roadmap', 'product', 'ops', 'workflow', 'issue', 'planning', 'productivity', '운영', '워크플로우', '프로덕트', '계획'],
+    tokens: {
+      bg: '#f7f8fb',
+      text: '#101114',
+      muted: '#5f6673',
+      primary: '#5e6ad2',
+      accent: '#26a69a',
+      surface: '#ffffff',
+      rule: '#dfe3ea',
+      display_px: 62,
+      body_px: 27,
+      caption_px: 14,
+      line_height: 1.38,
+      radius_px: 6,
+      treatment: 'precise_subtle_product_grid',
+      composition: 'operational_decision_matrix',
+      mono_label: 'compact status labels, dense but quiet operations layout'
+    },
+    applied_rules: [
+      'use a quiet operational canvas with dense hierarchy',
+      'keep the purple accent sparse and semantic',
+      'make comparison rows easy to scan',
+      'avoid marketing-style hero composition'
+    ]
+  }
+]);
+
 export const PPT_REQUIRED_GATE_FIELDS = Object.freeze([
   'clarification_contract_sealed',
   'audience_strategy_sealed',
@@ -282,6 +375,13 @@ export function buildPptStoryboard(contract = {}, audience = buildPptAudienceStr
 
 export function buildPptStyleTokens(contract = {}) {
   const korean = /[ㄱ-ㅎ가-힣]/.test(`${contract.prompt || ''} ${JSON.stringify(contract.answers || {})}`);
+  const reference = selectPptDesignReference(contract);
+  const refTokens = reference.applied_token_profile.color;
+  const fontStack = korean
+    ? '"Pretendard Variable", Pretendard, -apple-system, BlinkMacSystemFont, "Apple SD Gothic Neo", "Noto Sans KR", sans-serif'
+    : reference.primary.id.endsWith(':ibm')
+      ? '"IBM Plex Sans", -apple-system, BlinkMacSystemFont, "SF Pro Display", Inter, Arial, sans-serif'
+      : '-apple-system, BlinkMacSystemFont, "SF Pro Display", Inter, "Helvetica Neue", Arial, sans-serif';
   return {
     schema_version: 1,
     created_at: nowIso(),
@@ -294,23 +394,31 @@ export function buildPptStyleTokens(contract = {}) {
       gutter_px: 24
     },
     color: {
-      bg: '#f7f8fa',
-      text: '#111318',
-      muted: '#5b6270',
-      primary: '#0b5cff',
-      accent: '#00a88f',
-      surface: '#ffffff',
-      rule: '#d7dce5'
+      bg: refTokens.bg,
+      text: refTokens.text,
+      muted: refTokens.muted,
+      primary: refTokens.primary,
+      accent: refTokens.accent,
+      surface: refTokens.surface,
+      rule: refTokens.rule
     },
     typography: {
       language: korean ? 'ko' : 'en',
-      font_stack: korean
-        ? '"Pretendard Variable", Pretendard, -apple-system, BlinkMacSystemFont, "Apple SD Gothic Neo", "Noto Sans KR", sans-serif'
-        : '-apple-system, BlinkMacSystemFont, "SF Pro Display", Inter, "Helvetica Neue", Arial, sans-serif',
-      display_px: 76,
-      body_px: 30,
-      caption_px: 16,
-      line_height: korean ? 1.42 : 1.32
+      font_stack: fontStack,
+      display_px: refTokens.display_px,
+      body_px: refTokens.body_px,
+      caption_px: refTokens.caption_px,
+      line_height: korean ? Math.max(1.4, refTokens.line_height) : refTokens.line_height,
+      letter_spacing: 0
+    },
+    layout: {
+      composition: refTokens.composition,
+      treatment: refTokens.treatment,
+      radius_px: refTokens.radius_px,
+      rule_px: 1,
+      source_rail: true,
+      evidence_grid: true,
+      mono_label: refTokens.mono_label
     },
     design_policy: {
       priority: 'information_first',
@@ -328,8 +436,9 @@ export function buildPptStyleTokens(contract = {}) {
         authority: DESIGN_SYSTEM_SSOT.authority_file,
         builder_prompt: DESIGN_SYSTEM_SSOT.builder_prompt,
         route_local_artifact: PPT_STYLE_TOKENS_ARTIFACT,
-        rule: 'PPT style tokens are a route-local projection of the design SSOT; source inputs are fused here and are not independent authorities.'
+        rule: 'PPT style tokens are a route-local projection of the design SSOT; source inputs are selected, fused, and applied here rather than kept as independent authorities.'
       },
+      design_reference_selection: reference,
       source_inputs: [
         {
           id: GETDESIGN_REFERENCE.id,
@@ -343,31 +452,112 @@ export function buildPptStyleTokens(contract = {}) {
         }
       ],
       avoid: ['over-designed decoration', 'ornamental gradients', 'nested cards', 'low-contrast gray body text', 'excessive motion or effects'],
-      detail_strategy: ['precise spacing', 'clear hierarchy', 'thin rules', 'disciplined alignment', 'subtle accent color only when it clarifies meaning'],
-      anti_generic_ai_style: 'prevent AI-like design: select a concrete DESIGN.md or route-local visual system before styling; do not default to generic cards, gradients, vague SaaS visuals, oversized decoration, or unsupported image-like flourishes',
+      detail_strategy: ['precise spacing', 'clear hierarchy', 'thin rules', 'disciplined alignment', 'visible source rails', 'subtle accent color only when it clarifies meaning'],
+      anti_generic_ai_style: 'prevent AI-like design: select and apply a concrete awesome-design-md reference profile before styling; do not default to generic cards, gradients, vague SaaS visuals, oversized decoration, or unsupported image-like flourishes',
       image_policy: 'use images only when they improve comprehension; prefer Codex App built-in image generation via https://developers.openai.com/codex/app/features#image-generation when generated assets are needed'
     }
   };
 }
 
+export function selectPptDesignReference(contract = {}) {
+  const text = cleanText(`${contract.prompt || ''} ${JSON.stringify(contract.answers || {})}`).toLowerCase();
+  const scored = PPT_DESIGN_REFERENCE_PROFILES.map((profile) => {
+    const score = profile.keywords.reduce((sum, keyword) => sum + (text.includes(String(keyword).toLowerCase()) ? 1 : 0), 0);
+    return { profile, score };
+  }).sort((a, b) => b.score - a.score);
+  const primary = scored[0]?.score > 0 ? scored[0].profile : PPT_DESIGN_REFERENCE_PROFILES[0];
+  const secondary = scored.find((entry) => entry.profile.id !== primary.id && entry.score > 0)?.profile || PPT_DESIGN_REFERENCE_PROFILES.find((entry) => entry.id !== primary.id);
+  return {
+    source: AWESOME_DESIGN_MD_REFERENCE.url,
+    selection_method: 'keyword_match_against_sealed_ppt_contract',
+    primary: {
+      id: primary.id,
+      name: primary.name,
+      source_url: primary.source_url,
+      source_summary: primary.source_summary,
+      applied_rules: primary.applied_rules
+    },
+    secondary: secondary ? {
+      id: secondary.id,
+      name: secondary.name,
+      source_url: secondary.source_url,
+      source_summary: secondary.source_summary,
+      applied_rules: secondary.applied_rules.slice(0, 2)
+    } : null,
+    selected_sources: [primary, secondary].filter(Boolean).map((profile) => ({
+      id: profile.id,
+      name: profile.name,
+      source_url: profile.source_url,
+      role: profile.id === primary.id ? 'primary_style_reference' : 'secondary_guardrail_reference'
+    })),
+    applied_token_profile: {
+      color: primary.tokens,
+      composition: primary.tokens.composition,
+      treatment: primary.tokens.treatment
+    },
+    selection_reason: scored[0]?.score > 0
+      ? `matched ${scored[0].score} contract keyword(s) to ${primary.name}`
+      : `no strong domain match; defaulted to ${primary.name} for restrained business presentation output`
+  };
+}
+
 export function buildPptHtml({ contract = {}, audience, sourceLedger, storyboard, styleTokens }) {
   const title = escapeHtml(storyboard.title);
+  const referenceName = escapeHtml(styleTokens.design_policy?.design_reference_selection?.primary?.name || 'selected design reference');
+  const audienceRaw = escapeHtml(audience?.audience_profile?.raw || 'Audience context');
+  const stpRaw = escapeHtml(audience?.stp?.raw || 'STP context');
+  const decisionRaw = escapeHtml(audience?.decision_context?.raw || storyboard.thesis || '');
+  const surfaceRule = styleTokens.layout?.treatment === 'shadow_as_border_minimal_depth'
+    ? `box-shadow: 0 0 0 1px ${styleTokens.color.rule}; border: 0;`
+    : `border: 1px solid ${styleTokens.color.rule}; box-shadow: none;`;
   const css = `@page { size: 16in 9in; margin: 0; }
 * { box-sizing: border-box; }
 body { margin: 0; background: ${styleTokens.color.bg}; color: ${styleTokens.color.text}; font-family: ${styleTokens.typography.font_stack}; }
-.page { width: 100vw; min-height: 100vh; page-break-after: always; padding: 72px 96px; display: grid; align-content: center; gap: 26px; }
-.kicker { color: ${styleTokens.color.primary}; font-size: 18px; font-weight: 700; letter-spacing: 0; text-transform: uppercase; }
-h1 { margin: 0; font-size: 72px; line-height: 1.08; letter-spacing: 0; max-width: 1120px; }
-p { margin: 0; color: ${styleTokens.color.muted}; font-size: 28px; line-height: ${styleTokens.typography.line_height}; max-width: 920px; }
-.panel { border-left: 6px solid ${styleTokens.color.primary}; padding-left: 26px; }
-.source { font-size: 14px; color: ${styleTokens.color.muted}; align-self: end; }`;
+.page { width: 100vw; min-height: 100vh; page-break-after: always; padding: 64px 88px 54px; display: grid; grid-template-rows: auto 1fr auto; gap: 34px; }
+.topline { display: grid; grid-template-columns: 1fr auto; align-items: end; border-bottom: 1px solid ${styleTokens.color.rule}; padding-bottom: 14px; }
+.kicker { color: ${styleTokens.color.primary}; font-size: ${styleTokens.typography.caption_px}px; font-weight: 600; letter-spacing: 0; text-transform: uppercase; }
+.reference { color: ${styleTokens.color.muted}; font-size: ${styleTokens.typography.caption_px}px; letter-spacing: 0; }
+.content { display: grid; grid-template-columns: minmax(0, 6fr) minmax(320px, 4fr); gap: 58px; align-items: center; }
+h1 { margin: 0; font-size: ${styleTokens.typography.display_px}px; line-height: 1.08; letter-spacing: 0; max-width: 1040px; font-weight: 600; }
+p { margin: 0; color: ${styleTokens.color.muted}; font-size: ${styleTokens.typography.body_px}px; line-height: ${styleTokens.typography.line_height}; max-width: 920px; }
+.claim { display: grid; gap: 26px; }
+.evidence { ${surfaceRule} border-radius: ${styleTokens.layout.radius_px}px; background: ${styleTokens.color.surface}; display: grid; }
+.evidence-row { padding: 22px 24px; border-bottom: 1px solid ${styleTokens.color.rule}; }
+.evidence-row:last-child { border-bottom: 0; }
+.label { color: ${styleTokens.color.primary}; font-size: ${styleTokens.typography.caption_px}px; font-weight: 600; letter-spacing: 0; text-transform: uppercase; margin-bottom: 8px; }
+.value { color: ${styleTokens.color.text}; font-size: 20px; line-height: 1.42; }
+.source { display: grid; grid-template-columns: 1fr auto; gap: 24px; color: ${styleTokens.color.muted}; font-size: ${styleTokens.typography.caption_px}px; border-top: 1px solid ${styleTokens.color.rule}; padding-top: 14px; }
+.accent { width: 64px; height: 3px; background: ${styleTokens.color.accent}; }`;
   const pages = storyboard.pages.map((page) => `<section class="page">
-  <div class="kicker">${escapeHtml(page.kind)} / ${page.number}</div>
-  <div class="panel">
-    <h1>${escapeHtml(page.claim)}</h1>
-    <p>${escapeHtml(page.support)}</p>
+  <header class="topline">
+    <div class="kicker">${escapeHtml(page.kind)} / ${page.number}</div>
+    <div class="reference">${referenceName}</div>
+  </header>
+  <main class="content">
+    <div class="claim">
+      <div class="accent"></div>
+      <h1>${escapeHtml(page.claim)}</h1>
+      <p>${escapeHtml(page.support)}</p>
+    </div>
+    <aside class="evidence" aria-label="decision evidence">
+      <div class="evidence-row">
+        <div class="label">Audience</div>
+        <div class="value">${audienceRaw}</div>
+      </div>
+      <div class="evidence-row">
+        <div class="label">STP</div>
+        <div class="value">${stpRaw}</div>
+      </div>
+      <div class="evidence-row">
+        <div class="label">Decision</div>
+        <div class="value">${decisionRaw}</div>
+      </div>
+    </aside>
+  </main>
+  <div class="source">
+    <span>Sources: ${escapeHtml((page.source_ids || []).join(', ') || 'none')}</span>
+    <span>${escapeHtml(styleTokens.layout?.composition || 'presentation-grid')}</span>
   </div>
-  <div class="source">Sources: ${escapeHtml((page.source_ids || []).join(', ') || 'none')}</div>
 </section>`).join('\n');
   return `<!doctype html>
 <html lang="${styleTokens.typography.language}">
@@ -495,6 +685,9 @@ export function buildPptRenderReport({ contract = {}, audience, sourceLedger, st
       { id: 'restrained_detail', passed: styleTokens.design_policy?.visual_style === 'simple_restrained_detailed' },
       { id: 'design_ssot_declared', passed: styleTokens.design_policy?.design_ssot?.authority === DESIGN_SYSTEM_SSOT.authority_file },
       { id: 'curated_design_md_input_fused', passed: (styleTokens.design_policy?.source_inputs || []).some((entry) => entry.url === AWESOME_DESIGN_MD_REFERENCE.url && entry.role === 'source_input_for_ssot') },
+      { id: 'concrete_design_reference_selected', passed: Boolean(styleTokens.design_policy?.design_reference_selection?.primary?.id && styleTokens.design_policy?.design_reference_selection?.selected_sources?.length) },
+      { id: 'reference_rules_applied_to_tokens', passed: Boolean(styleTokens.layout?.composition && styleTokens.layout?.treatment && styleTokens.design_policy?.design_reference_selection?.applied_token_profile) },
+      { id: 'html_uses_reference_layout', passed: typeof html === 'string' && html.includes('decision evidence') && html.includes(styleTokens.layout?.composition || 'presentation-grid') },
       { id: 'ppt_skill_allowlist_enforced', passed: JSON.stringify(styleTokens.design_policy?.pipeline_allowlist?.required_skills || []) === JSON.stringify([...PPT_PIPELINE_SKILL_ALLOWLIST]) },
       { id: 'out_of_pipeline_design_skills_ignored', passed: styleTokens.design_policy?.pipeline_allowlist?.ignore_installed_out_of_pipeline_skills === true && (styleTokens.design_policy?.pipeline_allowlist?.ignored_design_skills_even_if_installed || []).includes('design-artifact-expert') },
       { id: 'ppt_mcp_allowlist_scoped', passed: (styleTokens.design_policy?.pipeline_allowlist?.allowed_mcp_servers || []).every((entry) => entry.mcp === 'context7' && /external_documentation/.test(entry.condition || '')) },

package/src/core/questions.mjs

@@ -131,6 +131,10 @@ function promptHasRisk(lower) {
   return /(운영|production|prod|live|배포|publish|release|결제|payment|billing|auth|인증|보안|security|db|database|supabase|postgres|sql|schema|migration|마이그레이션|삭제|delete|drop|truncate|reset|권한|permission|credential|secret)/.test(lower);
 }
 
+function promptNeedsExplicitRiskBoundary(lower) {
+  return /(drop|truncate|wipe|reset\s+(?:db|database)|delete\s+all|all-row|전체\s*(?:삭제|초기화)|운영\s*(?:db|데이터|삭제|쓰기|변경)|production\s*(?:db|database|delete|write|mutation)|prod\s*(?:db|database|delete|write|mutation)|credential|secret|api\s*key|토큰\s*(?:노출|삭제|교체)|권한\s*(?:상승|확대)|permission\s*(?:escalation|widening))/.test(lower);
+}
+
 function promptHasContextTarget(text, lower) {
   return promptHasTarget(text, lower)
     || /https?:\/\/\S+/.test(text)
@@ -146,8 +150,11 @@ export function buildAmbiguityAssessment(prompt, explicitAnswers = {}) {
   const acceptance = promptHasExplicitAcceptance(lower) || hasAnswer(explicitAnswers.ACCEPTANCE_CRITERIA) || hasAnswer(explicitAnswers.SUCCESS_CRITERIA_OR_ACCEPTANCE);
   const risk = promptHasRisk(lower);
   const contextTarget = promptHasContextTarget(text, lower) || hasAnswer(explicitAnswers.CODEBASE_CONTEXT_TARGET);
+  const actionable = target && action && !underspecified;
+  const hardRiskNeedsBoundary = promptNeedsExplicitRiskBoundary(lower);
   const predictableSafetyDefault = /(재시도|retry|세션\s*만료|session\s*expired|session\s*expiry|token\s*expired)/.test(lower);
-  const hasPolicy = hasAnswer(explicitAnswers.RISK_BOUNDARY) || hasAnswer(explicitAnswers.RISK_AND_BOUNDARY) || predictableSafetyDefault || /(하지\s*마|금지|no\s+|never|묻지|보존|preserve|safe|안전|검증|approval|승인|알아서|판단|추론|infer|default|기본)/.test(lower);
+  const predictableImplementationBoundary = actionable && risk && !hardRiskNeedsBoundary;
+  const hasPolicy = hasAnswer(explicitAnswers.RISK_BOUNDARY) || hasAnswer(explicitAnswers.RISK_AND_BOUNDARY) || predictableSafetyDefault || predictableImplementationBoundary || /(하지\s*마|금지|no\s+|never|묻지|보존|preserve|safe|안전|검증|approval|승인|알아서|판단|추론|infer|default|기본)/.test(lower);
   const hasMultipleChoiceRisk = /(\bor\b|또는|아니면|선택|둘 중|여러|multiple|대안)/.test(lower) && !/(알아서|판단|infer|추론|default|기본)/.test(lower);
 
   const goalClarity = underspecified ? (target || action ? 0.45 : 0.2) : (target && action ? 0.9 : target || action ? 0.62 : 0.25);
@@ -169,7 +176,7 @@ export function buildAmbiguityAssessment(prompt, explicitAnswers = {}) {
   const unresolved = [];
   if (components.goal.clarity_score < CLARITY_FLOORS.goal) unresolved.push('intent_target_or_required_outcome');
   if (components.success.clarity_score < CLARITY_FLOORS.success && (!target || !action || risk)) unresolved.push('success_criteria_or_acceptance');
-  if (components.constraints.clarity_score < CLARITY_FLOORS.constraints || hasMultipleChoiceRisk) unresolved.push('risk_boundary_or_choice');
+  if ((components.constraints.clarity_score < CLARITY_FLOORS.constraints && hardRiskNeedsBoundary) || hasMultipleChoiceRisk) unresolved.push('risk_boundary_or_choice');
   if (components.context.clarity_score < CLARITY_FLOORS.context) unresolved.push('codebase_context_target');
   const uniqueUnresolved = [...new Set(unresolved)];
   return {

package/src/core/routes.mjs

@@ -679,6 +679,7 @@ export function subagentExecutionPolicyText(route, prompt = '') {
   }
   return [
     'Subagent policy: REQUIRED for code-changing or execution work in this route.',
+    'The selected SKS route itself authorizes route-owned worker/reviewer subagents; the user does not need to separately ask for subagents when the default Team pipeline is active.',
     'Before editing, the parent orchestrator must visibly state the SKS route, split independent write scopes, and spawn worker/reviewer subagents whenever the tools are available.',
     'Run workers in parallel only with disjoint ownership. The parent owns integration, verification, and final evidence.',
     'If subagent tools are unavailable or the work cannot be safely split, record that as explicit subagent evidence before editing.',

package/src/core/tmux-ui.mjs

@@ -6,34 +6,113 @@ import { getCodexInfo } from './codex-adapter.mjs';
 import { codexAppIntegrationStatus, formatCodexAppStatus } from './codex-app.mjs';
 
 export const SKS_TMUX_LOGO = [
-  '   _____   __ __   _____',
-  '  / ___/  / //_/  / ___/',
-  '  \\__ \\  / ,<     \\__ \\ ',
-  ' ___/ / / /| |   ___/ / ',
-  '/____/ /_/ |_|  /____/  ',
-  '      SNEAKOSCOPE CODEX'
+  '        _______     __  __     _______',
+  '      / _____/|   / /_/ /|   / _____/|',
+  '     / /____| |  / __  / |  / /____| |',
+  '     \\____  \\ | / / / /| |  \\____  \\ |',
+  ' ____/ /  | |/_/ /_/ / | |____/ /  | |',
+  '/_____/   |//_/\\__/ /  |//_____/   |/',
+  '\\_____\\___/ \\_\\ \\_\\/___/ \\_____\\___/',
+  '          SNEAKOSCOPE CODEX'
 ].join('\n');
 
 const SKS_TMUX_LOGO_FRAMES = [
   [
-    '      __ __',
-    '     / //_/        .',
-    '    / ,<       .',
-    '   / /| |   .',
-    '  /_/ |_|',
-    '      S K S'
+    '             ||',
+    '             ||',
+    '             ||',
+    '             ||',
+    '             ||',
+    '             ||',
+    '             SKS',
+    '        SNEAKOSCOPE CODEX'
   ].join('\n'),
   [
-    '   _____   __ __   _____',
-    '  / ___/  / //_/  / ___/',
-    '  \\__ \\  / ,<     \\__ \\ ',
-    '     / / /| |      / /  ',
-    '    /_/ /_/ |_|   /_/   ',
-    '      SNEAKOSCOPE'
+    '          //||',
+    '        //  ||        .',
+    '       //   ||      .:',
+    '      //    ||    .::',
+    '     //     ||  .:::',
+    '    //      ||.::::',
+    '          S K S',
+    '        SNEAKOSCOPE CODEX'
+  ].join('\n'),
+  [
+    '        _______     __  __     _______',
+    '      / _____/|   / /_/ /|   / _____/|',
+    '     / /____| |  / __  / |  / /____| |',
+    '     \\____  \\ | / / / /| |  \\____  \\ |',
+    ' ____/ /  | |/_/ /_/ / | |____/ /  | |',
+    '/_____/   |//_/\\__/ /  |//_____/   |/',
+    '\\_____\\___/ \\_\\ \\_\\/___/ \\_____\\___/',
+    '          SNEAKOSCOPE CODEX'
+  ].join('\n'),
+  [
+    '        _______   __  __   _______',
+    '      / _____/  / /_/ /  / _____/|',
+    '     / /____   / __  /  / /____ | |',
+    '     \\____  \\ / / / /   \\____  \\| |',
+    ' ____/ /  /_/ /_/ /  ____/ /  | |',
+    '/_____/   /_/\\__/  /_____/   |/',
+    ' \\_____\\  \\_\\ \\_\\  \\_____\\___/',
+    '          SNEAKOSCOPE CODEX'
+  ].join('\n'),
+  [
+    '        _______     __  __     _______',
+    '      / _____/|   / /_/ /|   / _____/|',
+    '     / /____| |  / __  / |  / /____| |',
+    '     \\____  \\ | / / / /| |  \\____  \\ |',
+    ' ____/ /  | |/_/ /_/ / | |____/ /  | |',
+    '/_____/   |//_/\\__/ /  |//_____/   |/',
+    '\\_____\\___/ \\_\\ \\_\\/___/ \\_____\\___/',
+    '          SNEAKOSCOPE CODEX'
+  ].join('\n'),
+  [
+    '        _______   __  __   _______',
+    '      |\\_____ \\  / /_/ /  |\\_____ \\',
+    '      | |____\\ \\/ __  /   | |____\\ \\',
+    '      | |\\____\\/ / / /    | |\\____\\ \\',
+    '      | | |___/ /_/ /__   | | |___/ /',
+    '      \\|_|/____/\\__/__/   \\|_|/____/',
+    '           S K S',
+    '        SNEAKOSCOPE CODEX'
+  ].join('\n'),
+  [
+    '             ||\\\\',
+    '        .    ||  \\\\',
+    '       ::.   ||   \\\\',
+    '      ::::.  ||    \\\\',
+    '     ::::::. ||     \\\\',
+    '    :::::::::||      \\\\',
+    '          S K S',
+    '        SNEAKOSCOPE CODEX'
+  ].join('\n'),
+  [
+    '             ||',
+    '             ||',
+    '             ||',
+    '             ||',
+    '             ||',
+    '             ||',
+    '             SKS',
+    '        SNEAKOSCOPE CODEX'
   ].join('\n'),
   SKS_TMUX_LOGO
 ];
 
+const SKS_TMUX_LOGO_ANIMATION_STEPS = Object.freeze([
+  { frame: 0, color: '39', bold: false, delay: '0.045' },
+  { frame: 1, color: '39', bold: false, delay: '0.045' },
+  { frame: 2, color: '45', bold: false, delay: '0.05' },
+  { frame: 3, color: '51', bold: false, delay: '0.055' },
+  { frame: 4, color: '51', bold: true, delay: '0.07' },
+  { frame: 5, color: '51', bold: true, delay: '0.07' },
+  { frame: 6, color: '45', bold: false, delay: '0.05' },
+  { frame: 7, color: '39', bold: false, delay: '0.045' },
+  { frame: 8, color: '39', bold: false, delay: '0.045' },
+  { frame: 9, color: '51', bold: true, delay: '0.16' }
+]);
+
 export const DEFAULT_SKS_CODEX_MODEL = 'gpt-5.5';
 export const DEFAULT_SKS_CODEX_REASONING = 'high';
 
@@ -123,7 +202,7 @@ export function tmuxStatusKind(tmux = {}) {
 export function codexLaunchCommand(root, codexBin, codexArgs = []) {
   const extraArgs = Array.isArray(codexArgs) ? codexArgs : [];
   return [
-    sksLogoIntroCommand(),
+    sksLogoIntroCommand(codexBin),
     `printf '\\nProject: %s\\n' ${shellEscape(root)}`,
     'printf \'Runtime: tmux session for Codex CLI\\n\'',
     'printf \'Prompt:  use canonical $ commands, for example $Team or $QA-LOOP\\n\\n\'',
@@ -133,20 +212,25 @@ export function codexLaunchCommand(root, codexBin, codexArgs = []) {
   ].join('; ');
 }
 
-export function sksLogoIntroCommand() {
+export function sksLogoIntroCommand(codexBin = 'codex') {
   const staticLogo = `clear; printf '\\033[1;38;5;51m%s\\033[0m\\n' ${shellEscape(SKS_TMUX_LOGO)}`;
+  const authenticatedCheck = `${shellEscape(codexBin)} login status >/dev/null 2>&1`;
   const animated = [
     'clear',
-    `printf '\\033[38;5;39m%s\\033[0m\\n' ${shellEscape(SKS_TMUX_LOGO_FRAMES[0])}`,
-    'sleep 0.07',
-    'clear',
-    `printf '\\033[1;38;5;45m%s\\033[0m\\n' ${shellEscape(SKS_TMUX_LOGO_FRAMES[1])}`,
-    'sleep 0.08',
-    'clear',
-    `printf '\\033[1;38;5;51m%s\\033[0m\\n' ${shellEscape(SKS_TMUX_LOGO_FRAMES[2])}`,
-    'sleep 0.15'
+    'trap \'printf "\\033[0m\\033[?25h"\' EXIT INT TERM',
+    `printf '\\033[?25l'`,
+    ...SKS_TMUX_LOGO_ANIMATION_STEPS.flatMap((step) => {
+      const style = `${step.bold ? '1;' : ''}38;5;${step.color}`;
+      return [
+        `printf '\\033[H\\033[J\\033[${style}m%s\\033[0m\\n' ${shellEscape(SKS_TMUX_LOGO_FRAMES[step.frame])}`,
+        `sleep ${step.delay}`
+      ];
+    }),
+    `printf '\\033[H\\033[J\\033[1;38;5;51m%s\\033[0m\\n' ${shellEscape(SKS_TMUX_LOGO)}`,
+    `printf '\\033[?25h'`,
+    'trap - EXIT INT TERM'
   ].join('; ');
-  return `if [ "\${SKS_TMUX_LOGO_ANIMATION:-1}" = "0" ]; then ${staticLogo}; else ${animated}; fi`;
+  return `if [ -n "\${TMUX:-}" ] || [ "\${SKS_TMUX_LOGO_ANIMATION:-1}" = "0" ] || ${authenticatedCheck}; then ${staticLogo}; else ${animated}; fi`;
 }
 
 function terminalTitleCommand(title = '') {