lazycodex-ai 5.0.0-beta.4 → 5.0.0-beta.42

package/packages/shared-skills/skills/ultimate-browsing/ATTRIBUTION.md

@@ -1,19 +1,46 @@
 # ATTRIBUTION / NOTICE
 
 This skill (`ultimate-browsing`, part of `@oh-my-opencode/shared-skills`) ships
-project-original content plus two third-party tools that it installs at runtime
-(it does NOT vendor their source). Each runtime tool's license and required
-notices are reproduced below.
+project-original content, one vendored-and-modified upstream engine, plus two
+third-party tools that it installs at runtime (it does NOT vendor their source).
+Each component's provenance, license, and required notices are reproduced below.
 
 ---
 
-## 1. Project-original content (no third-party source vendored)
+## 1. insane-search engine — vendored upstream snapshot, modified
+
+`engine/**` originates from the **insane-search** project and is NOT
+project-original code, despite being heavily modified since import.
+
+- Upstream source: https://github.com/fivetaku/insane-search
+- Vendored into this repository on 2026-06-21 by commit
+  **`a4e4ed797`** (`feat(ultimate-browsing): vendor insane-search engine (junk-excluded)`),
+  via an explicit file whitelist that excluded caches and smoke-test junk.
+- Baseline: the upstream state as of that date, a **pre-0.7.0 snapshot**.
+  Upstream's CHANGELOG dates 0.7.0 to 2026-06-22; imported files carry no
+  version marker. We have never re-vendored since; the tree has diverged in
+  both directions.
+- Modifications by this project (non-exhaustive): de-personalization
+  (`4743199a5`), the Phase 2.5 surrogate retrieval stage and surrogate registry,
+  the provenance/trust result contract, the `bias_check.py` no-site-name CI gate,
+  module split of the fetch chain, and the Python test suite under
+  `engine/tests/`.
+- No upstream `LICENSE` file was included in the vendored snapshot, so this
+  repository has no upstream license text to reproduce here. Do not infer
+  project-original licensing from that absence; treat `engine/**` as
+  upstream-derived when reasoning about provenance.
+
+The binding version policy — which upstream baseline we sit on, why we stay
+pinned, what a future re-vendor must preserve, and what it must not import — is
+[`engine/AGENTS.md` §UPSTREAM BASELINE AND VERSION POLICY](engine/AGENTS.md).
+
+---
+
+## 2. Project-original content (no third-party source vendored)
 
 The following are authored by the oh-my-openagent project and carry no third-party
 license obligation:
 
-- `engine/**` — the insane-search Tier-1 fetch engine (curl_cffi grid, WAF
-  detection, Playwright fallback templates, `bias_check.py` no-site-name gate).
 - `references/insane-search/**` and `references/agent-reach/**` — the Tier-1 and
   Tier-1.5 reference docs.
 - `scripts/extract_cookies.py`, `scripts/cookie_paths.py`, `scripts/cookie_crypto.py`
@@ -26,13 +53,13 @@ user installs separately; this skill includes none of their source.
 
 ---
 
-## 2. CloakBrowser (CloakHQ) — Tier-2 stealth Chromium (runtime dependency)
+## 3. CloakBrowser (CloakHQ) — Tier-2 stealth Chromium (runtime dependency)
 
 The Tier-2 stealth browser is **CloakBrowser**, installed at runtime via `pip`
 (`pip install cloakbrowser`). No CloakBrowser source is vendored in this repository.
 
 - Source: https://github.com/CloakHQ/CloakBrowser
-- Pinned runtime version: **0.4.10** (documented in `references/chrome-stealth.md`;
+- Pinned runtime version: **0.5.7** (documented in `references/chrome-stealth.md`;
   this is a documented version string, not an automated drift check).
 - Wrapper source license: MIT License.
 - Binary license: the compiled CloakBrowser Chromium binary downloaded by
@@ -73,13 +100,13 @@ SOFTWARE.
 
 ---
 
-## 3. agent-browser (vercel-labs) — Tier-2 CDP automation CLI (runtime dependency)
+## 4. agent-browser (vercel-labs) — Tier-2 CDP automation CLI (runtime dependency)
 
 The Tier-2 automation CLI is **agent-browser**, installed at runtime via `npm`
-(`npm i -g agent-browser`). No agent-browser source is vendored in this repository.
+(`bun add -g agent-browser`). No agent-browser source is vendored in this repository.
 
 - Source: https://github.com/vercel-labs/agent-browser
-- Pinned runtime version: **0.31.1** (documented in `references/chrome-stealth.md`;
+- Pinned runtime version: **0.34.0** (documented in `references/chrome-stealth.md`;
   documented version string, no automated drift check).
 - Licensed under the Apache License, Version 2.0 (the "License"); you may not use
   these files except in compliance with the License. You may obtain a copy of the

package/packages/shared-skills/skills/ultimate-browsing/SKILL.md

@@ -1,13 +1,13 @@
 ---
 name: ultimate-browsing
-description: "Escalation skill for blocked or hard-to-reach web access — load it when a normal browse/fetch is blocked (WAF, 403, Cloudflare, JS-only render, login-gated, or a platform a generic fetcher cannot read). Tiered router: TIER 1 insane-search (headless extraction + WAF bypass via curl_cffi TLS impersonation, yt-dlp, Jina Reader, public APIs, Playwright real-Chrome fallback); TIER 1.5 agent-reach (platform-native readers for Chinese and social platforms: Xiaohongshu, Douyin, Weibo, Bilibili, V2EX, WeChat, plus Twitter/Reddit/LinkedIn/GitHub); TIER 2 Chrome stealth (CloakBrowser stealth Chromium + agent-browser CDP for clicks, forms, screenshots, video, cookie login). Triggers: blocked site, bypass bot detection, cloudflare/WAF bypass, scrape, stealth browser, import cookies, fill form, screenshot, play youtube, xiaohongshu, douyin, weibo, bilibili, v2ex, wechat article, podcast transcript. NOT for simple searches (use web-search) or plain fetches (use webfetch)."
+description: "Reaches web pages a plain fetch cannot: renders JS, drives clicks and forms, captures screenshots, holds a login, and gets past WAF blocks through platform-native readers and stealth Chrome. Use for any page work beyond retrieving static text."
 ---
 
 # Ultimate Browsing
 
-Escalation web access for tasks a normal browse or fetch cannot complete. Reach for this skill the moment a page is blocked (WAF / 403 / Cloudflare), needs JS rendering, hides behind a login, or lives on a platform a generic fetcher cannot read. Escalate only when the cheaper tier cannot do the job:
+Web access for everything a plain fetch cannot finish: a page that renders in JS, a click or a form, a screenshot, a login that must persist across pages, or a host that blocks generic fetchers (WAF / 403 / Cloudflare). Start at the cheapest tier that can do the job and climb only when it cannot:
 
-**Tier 1 — insane-search** (headless extraction + WAF bypass) -> **Tier 1.5 — agent-reach** (platform-native APIs, esp. Chinese platforms) -> **Tier 2 — Chrome stealth** (real interaction via CloakBrowser + agent-browser).
+**Tier 1 — insane-search** (headless extraction + WAF bypass) -> **Tier 1.5 — agent-reach** (platform-native APIs, esp. Chinese platforms) -> **Tier 2 — a real browser**: 2a a code-driven kernel browser, 2b Chrome stealth (CloakBrowser + agent-browser) when the page fights back.
 
 ## PHASE 0 — ROUTE FIRST (MANDATORY)
 
@@ -23,11 +23,11 @@ User request
   +- podcast transcript / stock forum ----------------- TIER 1.5 agent-reach
   +- Twitter feed / LinkedIn profile / GitHub via CLI - TIER 1.5 agent-reach
   |
-  +- Tier 1/1.5 returned empty or partial ------------- TIER 2  Chrome stealth
-  +- click / fill form / scroll / interact ------------ TIER 2  Chrome stealth
-  +- screenshot / render / play video ----------------- TIER 2  Chrome stealth
-  +- login session across pages / inject cookies ------ TIER 2  Chrome stealth
-  +- test web app / QA / dogfood ---------------------- TIER 2  Chrome stealth
+  +- Tier 1/1.5 returned empty or partial ------------- TIER 2  2a kernel browser -> 2b stealth
+  +- click / fill form / scroll / interact ------------ TIER 2  2a kernel browser -> 2b stealth
+  +- screenshot / render / play video ----------------- TIER 2  2a kernel browser -> 2b stealth
+  +- login session across pages / inject cookies ------ TIER 2  2b Chrome stealth (profile + cookies)
+  +- test web app / QA / dogfood ---------------------- TIER 2  2a kernel browser -> 2b stealth
   |
   +- simple search query ------------------------------ NOT this skill (use web-search)
 ```
@@ -80,12 +80,29 @@ curl -s "https://www.v2ex.com/api/topics/hot.json"            # V2EX public API
 
 Routing table, per-platform auth (set `TWITTER_*` env vars, `gh auth login`, a transcription key — only if you have access), rate-limit notes, and known version quirks are in [references/agent-reach/README.md](references/agent-reach/README.md).
 
-## Tier 2 — Chrome stealth (real interaction)
+## Tier 2 — a real browser (real interaction)
 
 **When**: real interaction is needed (clicks, forms, screenshots, video, persistent login), or Tier 1/1.5 failed.
 
+### Tier 2a — kernel browser (default)
+
+Drive the page from the code cell you are already in, with no CLI process and no open CDP port. On a Bun >= 1.4 runtime that is `new Bun.WebView()` (`navigate`, `click`, `type`, `evaluate`, `screenshot`, raw `cdp`); elsewhere it is `playwright-core`/`puppeteer-core` against the local Chrome. Snapshots and screenshots come back in-process, so this is the cheapest way to answer "what does the page actually render".
+
+```js
+await using view = new Bun.WebView({ width: 1280, height: 800 })
+await view.navigate(url)
+const title = await view.evaluate("document.title")
+await view.screenshot({ path: "/tmp/page.png" })
+```
+
+Climb to 2b when the site detects automation (Turnstile, FingerprintJS, "unusual traffic"), when the flow needs a persistent logged-in profile or injected cookies, or when the kernel browser cannot reach the page at all.
+
+### Tier 2b — Chrome stealth (blocked or logged-in pages)
+
 CloakBrowser is a stealth Chromium with source-level fingerprint patches that passes Cloudflare Turnstile, FingerprintJS, BrowserScan, and 30+ detectors; agent-browser is the CDP automation CLI that drives it. Both are runtime-installed tools (not vendored here). Full setup, version pins, launch flow, cookie login, and cross-platform notes are in [references/chrome-stealth.md](references/chrome-stealth.md).
 
+NEVER clear cookies, cache, or site data (`Network.clearBrowserCookies`, `Storage.clearCookies`, `chrome.browsingData.remove`, "clear browsing data") on the user's real/main browser profile — it wipes their logged-in state everywhere. If the task needs that profile's login state, clone the profile directory first (`rsync -a <profile>/ <tmp-clone>/`) and launch CloakBrowser / agent-browser with the clone as the user-data-dir; run any clearing on the clone only.
+
 ```bash
 # 1. Launch CloakBrowser with CDP on :9242 (see chrome-stealth.md for install + venv).
 # 2. CloakBrowser launches tabless — open the first tab via CDP before any agent-browser command:

package/packages/shared-skills/skills/ultimate-browsing/engine/AGENTS.md

@@ -2,6 +2,116 @@
 
 **Generated:** 2026-08-10 / 38d268995
 
+## UPSTREAM BASELINE AND VERSION POLICY
+
+**READ THIS BEFORE TOUCHING `engine/**` OR PROPOSING AN UPSTREAM SYNC.**
+
+This engine is NOT project-original code. It is a vendored-and-modified snapshot of
+[fivetaku/insane-search](https://github.com/fivetaku/insane-search), and the version
+we run on is a deliberate choice, not an accident of neglect.
+
+### The pin
+
+| Fact | Value |
+|---|---|
+| Upstream project | `https://github.com/fivetaku/insane-search` |
+| Vendoring commit | `a4e4ed797` (2026-06-21) `feat(ultimate-browsing): vendor insane-search engine (junk-excluded)` |
+| De-personalization | `4743199a5` (2026-06-21) |
+| Pinned upstream baseline | upstream state as of 2026-06-21, **pre-0.7.0** (0.7.0 is dated 2026-06-22) |
+| Re-vendors since | none — every later change here is ours |
+
+We intentionally track a **pinned baseline plus local divergence**, not upstream HEAD.
+There is no submodule and no automated drift check for this engine (unlike the
+`frontend` skill's upstream submodules): the vendored files ARE the source of truth,
+and upstream is a reference we port FROM, deliberately, file by file.
+
+### Why we do not blind-rebase onto upstream HEAD
+
+1. **Upstream reset its public history.** On 2026-08-06 upstream published a single
+   squashed commit, `019ee16 refactor: reset public history at 0.14.0`, discarding the
+   prior public history through 0.13.x. There is no upstream commit graph to rebase onto and no
+   way to cherry-pick an individual upstream change by sha — only whole-file diffing
+   against a moving HEAD.
+2. **Upstream 0.14.0 REMOVED capability.** The endpoint-mining / internal-API
+   auto-derivation / site-recipe subsystems that upstream carried publicly between
+   0.12.0 and 0.14.0 are gone from upstream HEAD. Syncing to HEAD is therefore not
+   strictly an upgrade: parts of it are a downgrade relative to the intermediate
+   versions, and none of it is recoverable from the reset history.
+3. **Our tree diverged on purpose.** The KEEP list below is functionality upstream
+   never had. A wholesale overwrite with upstream HEAD would silently delete it.
+4. **Different threat model.** Our engine ships inside a published npm package and a
+   public marketplace mirror, under a CI no-site-name gate and a de-personalization
+   deny-list. Upstream carries neither constraint, so upstream code is not
+   drop-in-shippable here.
+
+### KEEP — our divergences a re-vendor MUST NOT regress
+
+These exist only in our tree. Any upstream sync that removes or bypasses one of them
+is a regression, not an upgrade:
+
+- **Phase 2.5 surrogate retrieval** (`surrogate.py`, `surrogates.yaml`) — archive /
+  reader / proxy routes tried before paying for a browser spin-up, with per-entry
+  `last_verified` staleness handling and `--allow-proxy` gating.
+- **Provenance / trust contract** (`result_schema.py`) — `Provenance` and `Trust`
+  literals on every result, so a snapshot can never be reported as the live page.
+- **Surrogate dead-end validation** (L1.5 in `validators.py`) — interstitial titles and
+  AMP-style redirect stubs rejected instead of returned as content.
+- **The no-site-name rule and its CI gate** (`bias_check.py`) — zero hard-coded site
+  names, brands, or target domains in `engine/**`.
+- **Module split of the fetch chain** — `curl_probe` / `referers` / `url_transforms` /
+  `waf_detector` / `validators` / `executor` / `summary` as separate modules rather than
+  one monolith.
+- **The Python test suite** under `engine/tests/` with its HTML/JSON fixtures.
+- **De-personalization** — no personal absolute paths, no personal auth token literals,
+  no personal browser choice; enforced by `depersonalization-gate.test.ts`.
+- **Skill-level layering** — the engine is Tier 1 under a router that also owns Tier 1.5
+  (agent-reach) and Tier 2 (CloakBrowser + agent-browser). Upstream has no such tiering.
+
+### WANT — upstream improvements worth porting forward
+
+Our snapshot predates these; they are wanted, and each must be ported as a reviewed,
+site-agnostic change that preserves every KEEP item above. Port individually; never as
+a tree overwrite:
+
+- **Content quality**: dedicated markdown conversion of fetched HTML, main-content
+  extraction, PDF text extraction, and JSON-LD rescue when the HTML body is thin.
+- **Transient-failure retry** and **render-merge** of statically fetched HTML with the
+  browser-rendered DOM.
+- **Differential block classification** — distinguishing a bot-detection block from an
+  infrastructure or authentication failure, instead of collapsing both into `challenge`.
+- **Additional stealth fetch backends** beyond the current Playwright templates, and
+  additional WAF vendor profiles.
+- **Per-host route learning** — remembering which route succeeded for a host, with a TTL
+  and a bounded store. Must stay runtime state, never committed site knowledge (R4).
+- **Engine-level Phase 0 routing** — the official-public-API preference is currently only
+  a documented rule (R5) the agent can skip; upstream moved it into code so it cannot be
+  skipped. Worth adopting.
+
+### OUT OF SCOPE
+
+- **The removed upstream endpoint-mining / internal-API auto-derivation / site-recipe
+  subsystems.** They are absent from upstream HEAD and are not reconstructed here. They
+  also sit against R3/R4 and R7's anti-bias rule: discovered internal endpoints are
+  runtime findings, never committed engine knowledge.
+- **Any upstream code carrying site-specific selectors, domains, or brand names.** It
+  fails `bias_check.py` at the door; re-derive it site-agnostically or leave it out.
+- **Automated upstream tracking.** No submodule, no drift check, no auto-bump. Syncing is
+  a deliberate, reviewed, human-initiated act.
+
+### THE SYNC RULE
+
+Any future upstream sync preserves BOTH sides. Concretely:
+
+1. Diff the specific upstream capability you want against our tree — do not overwrite
+   files wholesale, and never `git checkout` upstream over `engine/`.
+2. Port it as its own reviewed change, keeping every KEEP item intact.
+3. Re-run `python3 engine/bias_check.py` and the `engine/tests/` suite; a port that
+   introduces a site name or breaks a fixture does not ship.
+4. Update the pin table above (baseline, date, what was ported) in the same change, plus
+   the provenance section of [`../ATTRIBUTION.md`](../ATTRIBUTION.md).
+5. If a port must drop a KEEP item, say so explicitly in the PR and get it agreed first —
+   silent regressions of the KEEP list are the failure mode this policy exists to prevent.
+
 ## OVERVIEW
 
 A 17-module Python package embedded in the `ultimate-browsing` skill: a site-agnostic fetch chain that escalates from a cheap curl probe to a real browser, with declarative WAF and surrogate registries. Not "optional scripts" — it has its own CLI entry (`python3 -m engine URL`), two YAML config schemas, a 4-file test suite, and a standalone CI guard. Package exports (`__init__.py`): `fetch`, `FetchResult`, `Attempt`, `Verdict`, `ValidationResult`, `validate`, `CHALLENGE_MARKERS`, `detect`, `TRANSFORMS`, `apply_transform`.

package/packages/shared-skills/skills/ultimate-browsing/engine/templates/package.json

@@ -4,7 +4,7 @@
   "private": true,
   "description": "Local deps for Playwright real-Chrome templates. npm install && npx playwright install chrome",
   "dependencies": {
-    "playwright": "^1.61.1",
+    "playwright": "^1.62.1",
     "playwright-extra": "^4.3.6",
     "puppeteer-extra-plugin-stealth": "^2.11.2"
   }

package/packages/shared-skills/skills/ultimate-browsing/references/chrome-stealth.md

@@ -2,8 +2,8 @@
 
 Real interaction (clicks, forms, screenshots, video, persistent login) for pages that defeat Tier 1/1.5. Two runtime tools, both installed on demand — neither is vendored in this skill:
 
-- **CloakBrowser** (`pip`) — stealth Chromium with source-level C++ fingerprint patches. The Python wrapper source is MIT; the downloaded Chromium binary is covered by CloakBrowser's separate binary license and is not redistributed by this package. Passes Cloudflare Turnstile, FingerprintJS, BrowserScan, and 30+ detectors. Pin **0.5.5**.
-- **agent-browser** (`npm`, Apache-2.0) — native CDP automation CLI that drives CloakBrowser. AX-tree snapshots, `@eN` refs, click/fill/type/scroll, screenshots, video, cookie/state/session management. Pin **0.33.2**.
+- **CloakBrowser** (`pip`) — stealth Chromium with source-level C++ fingerprint patches. The Python wrapper source is MIT; the downloaded Chromium binary is covered by CloakBrowser's separate binary license and is not redistributed by this package. Passes Cloudflare Turnstile, FingerprintJS, BrowserScan, and 30+ detectors. Pin **0.5.7**.
+- **agent-browser** (`npm`, Apache-2.0) — native CDP automation CLI that drives CloakBrowser. AX-tree snapshots, `@eN` refs, click/fill/type/scroll, screenshots, video, cookie/state/session management. Pin **0.34.0**.
 
 ```
 CloakBrowser (stealth Chromium) <- CDP port 9242 -> agent-browser CLI
@@ -17,30 +17,35 @@ CloakBrowser (stealth Chromium) <- CDP port 9242 -> agent-browser CLI
 
 CloakBrowser runs in a dedicated Python venv. Cross-platform: macOS, Linux, and Windows all supported by both tools (use the venv path convention for your OS).
 
+The venv lives at a fixed path so every session reuses one stealth-Chromium download instead of creating a new venv per working directory, and every command below calls its interpreter by absolute path — no `activate` step to forget.
+
 ```bash
-# CloakBrowser (MIT wrapper source; separate binary license, pin 0.5.5):
-uv venv .cloak-venv --python 3.13
-# macOS/Linux: source .cloak-venv/bin/activate    Windows: .cloak-venv\Scripts\activate
-uv pip install "cloakbrowser==0.5.5"
-python -c "import cloakbrowser; cloakbrowser.ensure_binary()"   # downloads stealth Chromium on first import
-
-# agent-browser (Apache-2.0, pin 0.33.2):
-npm i -g agent-browser@0.33.2 && agent-browser install
-agent-browser --version   # 0.33.2
+# CloakBrowser (MIT wrapper source; separate binary license, pin 0.5.7):
+VENV="$HOME/.agents/cloak-venv"          # Windows: %USERPROFILE%\.agents\cloak-venv
+uv venv "$VENV" --python 3.13
+uv pip install --python "$VENV/bin/python" "cloakbrowser==0.5.7"
+"$VENV/bin/python" -c "import cloakbrowser; cloakbrowser.ensure_binary()"   # downloads stealth Chromium on first import
+
+# agent-browser (Apache-2.0, pin 0.34.0):
+bun add -g agent-browser@0.34.0 && agent-browser install
+agent-browser --version   # 0.34.0
 ```
 
 Verify CloakBrowser:
 
 ```bash
-python -c "import cloakbrowser; print(cloakbrowser.__version__, cloakbrowser.CHROMIUM_VERSION, cloakbrowser.binary_info()['installed'])"
-# -> 0.5.5  <chromium-version>  True
+"$VENV/bin/python" -c "import cloakbrowser; print(cloakbrowser.__version__, cloakbrowser.CHROMIUM_VERSION, cloakbrowser.binary_info()['installed'])"
+# -> 0.5.7  <chromium-version>  True
 ```
 
 ## Launch + drive
 
+NEVER clear cookies, cache, or site data (`Network.clearBrowserCookies`, `Storage.clearCookies`, `chrome.browsingData.remove`, "clear browsing data") on the user's real/main browser profile — it wipes their logged-in state everywhere. If you need that profile's login state, clone it first (`rsync -a <profile>/ <tmp-clone>/`) and launch with the clone as the user-data-dir; run any clearing on the clone only.
+
 ```bash
-# 1. Launch CloakBrowser with CDP on :9242 (background). With the venv active:
-python -c "import asyncio,cloakbrowser; asyncio.run(cloakbrowser.launch_async(headless=False, stealth_args=True, args=['--remote-debugging-port=9242']))" &
+# 1. Launch CloakBrowser with CDP on :9242 (background). cloakbrowser lives only in the venv,
+#    so call its interpreter by absolute path — this works in any shell, activated or not:
+"$HOME/.agents/cloak-venv/bin/python" -c "import asyncio,cloakbrowser; asyncio.run(cloakbrowser.launch_async(headless=False, stealth_args=True, args=['--remote-debugging-port=9242']))" &
 
 # 2. CloakBrowser launches tabless -> agent-browser would say "No page found".
 #    Open the first tab via CDP before any agent-browser command:
@@ -76,7 +81,7 @@ agent-browser skills list                # everything available on the installed
 agent-browser --cdp 9242 eval 'navigator.webdriver'   # must print false
 ```
 
-Verified 2026-07 with CloakBrowser 0.5.5 + agent-browser 0.33.2: `navigator.webdriver` reads the boolean false with no init-script, bot.sannysoft.com all-green, browserscan.net "Normal" (15/15), nowsecure.nl Turnstile bypassed.
+Verified 2026-07 with CloakBrowser 0.5.7 + agent-browser 0.34.0: `navigator.webdriver` reads the boolean false with no init-script, bot.sannysoft.com all-green, browserscan.net "Normal" (15/15), nowsecure.nl Turnstile bypassed.
 
 > **agent-browser 0.33.x behavior note:** the daemon now defaults to a 1-hour idle timeout (saves restore state, closes the browser, exits after 1 h of no commands). Set `AGENT_BROWSER_IDLE_TIMEOUT_MS=0` to restore the old always-persist behavior. External WebSocket stream consumers see latest-wins frame delivery; `record` (CDP) and the dashboard are unaffected.
 
@@ -117,6 +122,6 @@ lsof -ti:9242 | xargs kill -9
 # agent-browser can't connect:
 curl -s http://127.0.0.1:9242/json/version | head -5   # empty -> CloakBrowser not running
 # Update either tool:
-uv pip install --upgrade "cloakbrowser==0.5.5" && python -c "import cloakbrowser; cloakbrowser.ensure_binary()"
-npm i -g agent-browser@0.33.2
+uv pip install --python "$HOME/.agents/cloak-venv/bin/python" --upgrade "cloakbrowser==0.5.7" && "$HOME/.agents/cloak-venv/bin/python" -c "import cloakbrowser; cloakbrowser.ensure_binary()"
+bun add -g agent-browser@0.34.0
 ```

package/packages/shared-skills/skills/ultimate-browsing/references/insane-search/README.md

@@ -16,7 +16,8 @@
 1. WebFetch, 즉흥 curl, 수동 헤더 조합 **시도 금지**
 2. 즉시 다음을 실행:
    ```bash
-   python3 -m engine "<URL>" [--selector "<CSS>"] [--device auto|desktop|mobile] [--trace]
+   python3 -m engine "<URL>" [--selector "<CSS>"] [--device auto|desktop|mobile] [--trace] \
+                              [--no-retry] [--no-extract]
    ```
 3. 종료코드 0(ok) 또는 1(fail) 받은 뒤 판단. trace를 먼저 읽고 재시도 결정.
 4. 실패 시에만 `--trace --json`으로 재호출해서 원인 진단 후 `--device` 또는 `user_hint` 조정.
@@ -135,7 +136,13 @@ result = fetch(
 
 if result.ok:
     print(result.verdict)     # strong_ok | weak_ok
-    html = result.content
+    html = result.content     # raw body — 단, content-rescue가 발동한 경우 구조 텍스트
+    # v0.10.0 content-rescue: PDF 응답은 pypdf 추출 텍스트, visible text가 얇은
+    # SPA 셸은 JSON-LD articleBody / 렌더된 innerText로 대체될 수 있다.
+    # result.extraction_source로 판별: "raw"(원문 그대로) | pdf | json_ld | *+inner_text.
+    # 일반 HTML 성공은 항상 raw. 끄기: enable_extraction=False / --no-extract.
+    # 429/502/503/504는 probe에서 백오프 재시도(Retry-After 반영, 총 10초 캡);
+    # 끄기: enable_retry=False / --no-retry.
 else:
     # Phase 3 수동 개입 (Playwright MCP) 필요 — result.trace로 원인 진단
     pass

package/packages/shared-skills/skills/{start-work → ulw-execute}/SKILL.md

@@ -1,6 +1,6 @@
 ---
-name: start-work
-description: "Execute a Prometheus work plan with Boulder state, evidence ledger updates, worktree discipline, parallel subagents, and Stop-hook continuation. Use after planning when the user says start work, execute plan, continue plan, resume plan, or asks to run a .omo/plans plan."
+name: ulw-execute
+description: "Executes a written Prometheus work plan with Boulder state, evidence ledger, worktree discipline, and parallel subagents. Use when the user says ulw-execute or asks to run a .omo/plans plan."
 ---
 
 ## ABSOLUTE RULE: YOU ARE AN ORCHESTRATOR — NEVER THE IMPLEMENTER
@@ -31,28 +31,28 @@ Every `multi_agent_v1.spawn_agent` message is a self-contained executable assign
 
 Plan and reviewer agents may run for a long time: spawn them in the background and keep doing independent root work. Between `multi_agent_v1.wait_agent` calls, back off — double the timeout up to ~5 minutes — instead of spinning short cycles. A timeout only means no new mailbox update arrived; treat a running child as alive. Require `WORKING: <task> - <current phase>` before long passes and `BLOCKED: <reason>` only when progress stops. Keep the parent visibly alive with active subagent count, names, and latest `WORKING:` phase. Fallback only when the child is completed without the deliverable, ack-only after followup, explicitly `BLOCKED:`, or no longer running — then record inconclusive (never a pass), close if safe, and respawn a smaller `fork_context: false` task with the missing deliverable.
 
-# start-work
+# ulw-execute
 
-Execute a Prometheus work plan until every top-level checkbox is complete. This skill pairs with the harness's start-work continuation hook, which re-injects the next turn while `.omo/boulder.json` says this `codex:<session_id>` still has unchecked plan work.
+Execute a Prometheus work plan until every top-level checkbox is complete. This skill pairs with the harness's ulw-execute continuation hook, which re-injects the next turn while `.omo/boulder.json` says this `codex:<session_id>` still has unchecked plan work.
 
 ## Usage
 
 ```text
-$start-work [plan-name] [--worktree <absolute-path>] [--make-pr] [--ship]
+$ulw-execute [plan-name] [--worktree <absolute-path>] [--make-pr] [--ship]
 ```
 
 - `plan-name` (optional): a full or partial file stem under `.omo/plans/`.
-- `--worktree` (required for PR/branch work; otherwise optional): the task-owned git worktree path.
-- `--make-pr` (optional): deliver the work as a pull request. IMPLIES worktree mode: when `--worktree` is absent, create a task-owned worktree (`git worktree add <absolute-path> <base-branch>`) before implementation and record it as `worktree_path`. On completion, push the branch and open a reviewer-readable PR, then hand off with the PR URL - merge only if the user asks.
+- `--worktree` (optional): reuse an existing task-owned worktree for the first phase instead of creating one; every phase runs in a task-owned worktree regardless.
+- `--make-pr` (optional): deliver each phase's worktree as a pull request — push the branch, open a reviewer-readable PR, hand off with the URL, and merge only if the user asks.
 - `--ship` (optional): full delivery lifecycle; implies `--make-pr`. After the PR opens, stay on the job until it is MERGED: watch CI and review gates, fix failures and address feedback from the worktree (fresh QA evidence for behavior changes), merge per the repository's merge policy, then remove the worktree and sync `.omo/` state back.
 
 ## Goal and todo discipline (MANDATORY)
 
 Do ALL of this immediately after the plan is selected, BEFORE the first implementation dispatch. Skipping any step is a defect.
 
-1. **Set the goal, in detail.** When a goal tool is available (`create_goal`), call it with a DETAILED objective: the plan name and path, the concrete end state, the phase and task counts, the delivery mode (direct, `--make-pr`, or `--ship`), and how completion will be verified. One work session = one goal. No goal tool -> record the same objective as the first ledger entry.
+1. **Set the goal, in detail.** When a goal tool is available (`create_goal`), call it with a DETAILED objective: the plan name and path, the concrete end state, the phase and task counts, the delivery mode (direct, `--make-pr`, or `--ship`), and how completion will be verified. One work session = one registered goal (the goal tool holds one active goal); each phase then carries its own concrete goal — the ledger entry Phase 2 records before the wave's first dispatch, defined from the previous phase's landed and verified evidence. No goal tool -> record the same objective as the first ledger entry.
 2. **Register every phase and task as todos.** Mirror the plan into the todo/plan tool of your harness: one phase per plan wave, one todo per column-zero checkbox (including the final verification wave). Register ALL of them up front - never keep tasks in memory only.
-3. **Keep them current at every moment.** Mark a todo in_progress when its work dispatches and done immediately after its verification passes. Never batch-complete at the end, never execute work that is not a registered todo; discovered work is appended as a todo before it runs. The todo list, Boulder state, and plan checkboxes must always tell the same story.
+3. **Keep them current at every moment.** Mark a todo in_progress when its work dispatches and done immediately after its verification passes. Never batch-complete at the end, never execute work that is not a registered todo; discovered work — a pre-existing bug, failing test, stale doc, or wrong guidance — is appended to the plan as a checkbox, mirrored as a todo before it runs, and fixed to the ideal state, never deferred as a follow-up. A worker that meets a defect outside its assigned files reports it instead of fixing it; only the orchestrator appends the checkbox and dispatches it to a correctly scoped unit. The todo list, Boulder state, and plan checkboxes must always tell the same story.
 
 ## Phase 1: Select the plan
 
@@ -66,7 +66,7 @@ Do ALL of this immediately after the plan is selected, BEFORE the first implemen
 
 ### No-plan bootstrap
 
-When the user explicitly said `start work` / `$start-work` and no selectable plan exists, treat that phrase as approval: bootstrap `ulw-plan` to create the approved plan before execution and implementation, instead of stalling or asking for generic approval again. A brief or notes file without waves, checkboxes, and acceptance criteria is NOT decision-complete — enter this bootstrap too.
+When the user explicitly said `start work` / `$ulw-execute` and no selectable plan exists, treat that phrase as approval: bootstrap `ulw-plan` to create the approved plan before execution and implementation, instead of stalling or asking for generic approval again. A brief or notes file without waves, checkboxes, and acceptance criteria is NOT decision-complete — enter this bootstrap too.
 
 1. Invoke the `ulw-plan` skill from the current request and require its dynamic adversarial workflow: collect, verify, design, adversarial plan-review, synthesize.
 2. The generated Prometheus plan must be saved under `.omo/plans/<slug>.md` before implementation or Boulder state writes that point at plan work.
@@ -95,13 +95,14 @@ Write `.omo/boulder.json` before implementation starts. Prefix session ids with
 }
 ```
 
-For PR/branch work, a task-owned worktree is mandatory before implementation starts: pass `--worktree`, or use `--make-pr`/`--ship`, which auto-create one. Verify the path with `git worktree list --porcelain` or create it with `git worktree add <path> <branch-or-HEAD>`, then store the absolute path as `worktree_path`. All edits, commands, tests, and evidence capture must run inside that worktree.
+Every phase (plan wave) runs in its own task-owned worktree with its own goal: before the wave's first dispatch, record the wave's goal — its checkboxes and their acceptance criteria — as a ledger entry, then `git worktree add <repo>-wt/<plan>-<wave> <branch off the integration base>` (or verify a `--worktree` path with `git worktree list --porcelain`), store the absolute path as `worktree_path`, run every edit, command, test, and evidence capture inside it; the wave lands on the integration base once its checkboxes are verified (direct merge, or the PR under `--make-pr`/`--ship`), and the next wave branches from that landed base.
 
 ## Parallel delivery lanes (teams and worktrees)
 
 Solo orchestration with parallel background workers is the default topology. Decide once, when the wave's lanes are known, and record the verdict in the ledger:
 
 - **Independent lanes -> parallel workers.** Separate files, no shared contract: one parallel spawn burst; no team.
+- **Dependency-ordered lanes -> one `workflow` run per wave.** Sub-tasks with real ordering between them (C needs A and B finished first) and a harness with a native `workflow` tool: dispatch the wave as ONE run (one producer node per lane plus a verification node); recover inside it with `retry`/`amend`/`send`; let node completions wake you instead of arming per-lane watchers; the next wave is a NEW run (or `amend` when only the definition changed) — never one graph for the whole plan. Read the `mass-ulw` skill's `SKILL.md` and `references/planning.md` IN FULL before defining any graph.
 - **Overlapping lanes -> a team.** The lanes touch the same module or contract AND running them concurrently actually finishes sooner: stand up a team (where the harness has one) so one lane's discoveries relay through you mid-flight.
 - **PR-mode independent lanes -> a worktree per lane.** Under `--make-pr`/`--ship`, when a wave holds independent checkboxes, give each lane its own branch and task-owned worktree, delivered as its own PR.
 
@@ -118,7 +119,7 @@ Landing rules, regardless of topology:
 3. Ignore nested checkboxes under acceptance criteria, evidence, and definition-of-done sections.
 4. Classify the checkbox tier and record it in its ledger entry. Default is LIGHT — a narrow change inside existing layers. Take HEAVY only on a fact you can point to: a new module / abstraction / domain model; auth, security, or session; an external integration; a DB schema or migration; concurrency or transaction boundaries; a cross-domain refactor; or the plan or user signals care. When unsure, take HEAVY; upgrade and redo skipped gates the moment a HEAVY fact surfaces; never downgrade.
 5. Decompose that checkbox into atomic sub-tasks sized for ONE worker in ONE run — a sub-task that would need mid-flight steering is two sub-tasks. Collect every other unchecked checkbox in the same plan wave whose dependencies are met — their lanes execute concurrently. A wave that could split further but holds fewer than 3 independent sub-tasks is under-split.
-6. **DELEGATE EVERYTHING. YOU NEVER IMPLEMENT.** Route every sub-task through the delegation router below, then dispatch ALL independent sub-tasks across those checkboxes in one parallel worker-spawn burst (a single batched spawn call where the harness supports it); serialize only named dependencies. Verification and checkbox marking stay per-checkbox.
+6. **DELEGATE EVERYTHING. YOU NEVER IMPLEMENT.** Route every sub-task through the delegation router below, then dispatch ALL independent sub-tasks across those checkboxes in one parallel worker-spawn burst (a single batched spawn call where the harness supports it); route named dependencies per the lane-topology decision above. Verification and checkbox marking stay per-checkbox.
 7. Give every dispatched sub-task its completion condition and watch for it per the section below. A dispatch whose completion nobody watches is an unfinished dispatch.
 
 ### Monitor every dispatched subagent to its completion condition
@@ -161,7 +162,7 @@ Each sub-task message must include:
    - Terminal / TUI: drive a real pty; `tmux send-keys` is fine for a boot/behavior smoke, but color/layout/CJK evidence goes through the xterm.js web terminal below, NEVER `tmux capture-pane`.
    - Browser use: prefer the harness's in-app browser control when available and the scenario does not need an authenticated or persistent user browser profile; otherwise drive the real page with Chrome, or agent-browser (https://github.com/vercel-labs/agent-browser) when Chrome is unavailable.
    - Computer use: OS-level GUI automation against the running desktop app when the surface is not a page.
-   - TUI visual evidence: when a TUI claim needs visual QA or PR proof, run `node script/qa/web-terminal-visual-qa.mjs --command "<cmd>" --input "{Enter}" --evidence-dir <dir>` (real pty rendered through xterm.js in Chrome) and attach `terminal.png` plus `metadata.json`.
+   - TUI visual evidence: when a TUI claim needs visual QA or PR proof, run `bun script/qa/web-terminal-visual-qa.mjs --command "<cmd>" --input "{Enter}" --evidence-dir <dir>` (real pty rendered through xterm.js in Chrome) and attach `terminal.png` plus `metadata.json`.
 6. The adversarial classes that apply to this sub-task (from the 9 ultraqa classes) and how each is probed.
 7. Required artifact path and cleanup receipt.
 8. Tool-use expectations: batch independent tool calls in parallel; when the harness exposes a code-execution surface (eval), use it for multi-call steps instead of one-by-one calls.
@@ -178,7 +179,7 @@ For each checkbox, complete all five gates before marking it done:
 4. Adversarial QA: exercise every class the Phase 3 trigger map marks applicable and capture the observable result for each.
 5. Cleanup: register every QA resource teardown as its own todo when spawned (QA scripts, tmux assets, browser sessions, PIDs, ports, containers, temp dirs), execute each, and capture the receipt. No QA asset is left running.
 
-Append evidence to `.omo/start-work/ledger.jsonl`, one JSON object per line. Include at least `event`, `plan`, `task`, `session_id`, `commands`, `artifact`, `adversarial_classes`, and `cleanup` fields. `adversarial_classes` lists each probed class with its observable result and each ruled-out class with a one-line reason.
+Append evidence to `.omo/ulw-execute/ledger.jsonl`, one JSON object per line. Include at least `event`, `plan`, `task`, `session_id`, `commands`, `artifact`, `adversarial_classes`, and `cleanup` fields. `adversarial_classes` lists each probed class with its observable result and each ruled-out class with a one-line reason.
 
 ### Sisyphus-style completion contract
 
@@ -224,7 +225,7 @@ Only after verification passes:
 When all top-level checkboxes in `## TODOs` and `## Final Verification Wave` are complete:
 
 1. Run the plan's final verification commands.
-2. For PR/branch work, finish the lifecycle from the task-owned worktree: sync `.omo/` state back to the main repo, create or update the PR, wait for review/verification gates, merge by default unless explicitly opted out, and remove the worktree only after successful merge or explicit handoff.
+2. For PR/branch work, finish the lifecycle from the last phase's worktree: sync `.omo/` state back to the main repo, create or update the PR, wait for review/verification gates, merge by default unless explicitly opted out, and remove the worktree only after successful merge or explicit handoff.
 3. Remove or mark the Boulder work as completed.
 4. Print an `ORCHESTRATION COMPLETE` block with the plan path, verification commands, artifacts, and cleanup receipts.
 
@@ -235,6 +236,6 @@ When all top-level checkboxes in `## TODOs` and `## Final Verification Wave` are
 - No tests-only completion claim. A Manual-QA artifact is required.
 - **NO DIRECT IMPLEMENTATION BY THE ORCHESTRATOR.** Root NEVER edits product files, writes tests, or runs QA itself — a spawned worker does.
 - No completion claim while an applicable ultraqa adversarial class was never probed. Each applicable class needs a captured observable result; each skipped class needs a one-line not-applicable reason in the ledger.
-- No PR/branch implementation, review, or merge in the main worktree; use the task-owned git worktree.
+- No implementation, review, or merge in the main checkout; every phase works in its task-owned worktree.
 - No unprefixed session ids in Boulder state. Sessions are always recorded as `codex:<session_id>`.
 - No stale-memory execution. The plan and ledger are the durable source of truth.

package/packages/shared-skills/skills/ulw-plan/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ulw-plan
-description: "ACTIVATES ONLY on an explicit user request for the ulw-plan workflow: the user themselves saying ulw-plan, ulw plan, /skill:ulw-plan, or asking in their own words for a work plan before coding. NEVER self-activates: a bare ulw/ultrawork run, an agent-side routing decision, or reading this file is not a request, and the plan-gated reviewers (metis/momus) stay locked without a user request plus a written .omo/plans plan file. Explore-first planning consultant (Prometheus) that grounds in the codebase, asks only the forks exploration cannot resolve - or researches them to best practice when the intent is fuzzy - waits for explicit approval, then writes ONE decision-complete work plan a worker executes with zero further interview. Triggers: ulw-plan, ulw plan, plan this, make a plan, plan before coding, interview me, break this down, start planning, plan mode."
+description: "Explore-first planning consultant that writes one decision-complete work plan before coding. Use only on an explicit user request for the ulw-plan workflow or a plan before implementation; it never self-activates on a bare ulw run."
 metadata:
   short-description: Explore-first planning consultant that waits for your okay before planning
 ---
@@ -9,7 +9,7 @@ metadata:
 
 You are **Prometheus**, a planning consultant. You turn a vague or large request into ONE **decision-complete** work plan a downstream worker executes with zero further interview. You read, search, run read-only analysis, and write ONLY plan artifacts under `.omo/`. You are a PLANNER - you never edit product code and never implement.
 
-**Plan mode is sticky.** "do X" / "fix X" / "build X" / "just do it" all mean "plan X". You **never start implementation** - not for small, obvious, or urgent work, and not through a subagent: delegated implementation is still implementation. Execution belongs to a separate worker session that only the user starts (e.g. `$start-work`).
+**Plan mode is sticky.** "do X" / "fix X" / "build X" / "just do it" all mean "plan X". You **never start implementation** - not for small, obvious, or urgent work, and not through a subagent: delegated implementation is still implementation. Execution belongs to a separate worker session that only the user starts (e.g. `$ulw-execute`).
 
 Outcome-first: explore a lot, ask few sharp questions - or none, when the intent is fuzzy (see routing) - and stop the moment the plan is done.
 
@@ -23,18 +23,18 @@ If another active mode mandates its own first line (ultrawork does), print that
 
 Directly under the marker, before any exploration, state the working contract once, in your own words, carrying ALL of these commitments:
 
-1. **Persona + no-implementation pledge** - from now on you work as Prometheus, a planning consultant, and you will never start implementation - no product-code edits, no implementer subagents - until the user explicitly says okay; even then, approval authorizes writing the plan only, and execution starts in a separate worker session (e.g. `$start-work`).
+1. **Persona + no-implementation pledge** - from now on you work as Prometheus, a planning consultant, and you will never start implementation - no product-code edits, no implementer subagents - until the user explicitly says okay; even then, approval authorizes writing the plan only, and execution starts in a separate worker session (e.g. `$ulw-execute`).
 2. **Workflow preview** - the order of what happens next: parallel read-only exploration (plus outside research when the repo cannot answer) until the open unknowns are resolved; the intent verdict from INTENT ROUTING, announced; questions to the user ONLY when a genuine owner-decision survives exploration - or when exploration and research both come back empty on a fork the plan cannot proceed without; then the approval brief, and the plan is written only after the explicit okay.
 
 Example opening (adapt the wording, keep every commitment):
 
 > ULW-PLAN MODE ENABLED!
-> From now on I am working as Prometheus, a planning consultant. I will not start any implementation until you explicitly say okay - and approval authorizes writing the plan only; execution starts separately (e.g. `$start-work`).
+> From now on I am working as Prometheus, a planning consultant. I will not start any implementation until you explicitly say okay - and approval authorizes writing the plan only; execution starts separately (e.g. `$ulw-execute`).
 > Next, in order: (1) parallel read-only exploration and research, (2) intent verdict announced (CLEAR or UNCLEAR, plus whether high-accuracy review is required), (3) questions only for the forks exploration cannot settle - or where research finds nothing on a blocking decision, (4) approval brief, then (5) the plan is written after your okay.
 
 ## INTENT ROUTING - pick ONE intent reference
 
-**Review modifiers are a gate trigger, not a style cue.** If the user says "high accuracy", "ultra high accuracy", "고정밀", "deep review", or equivalent - in ANY turn, even appended to a follow-up question and even after the plan already exists - set `review_required: true` in the draft: the dual high-accuracy review (native `momus` + the independent Oracle review) is now REQUIRED before handoff, and if the plan already exists you run it this same turn. Answering the current question more carefully does NOT satisfy it. This does NOT choose CLEAR/UNCLEAR and does NOT suppress interview.
+**Review modifiers are a gate trigger, not a style cue.** If the user says "high accuracy", "ultra high accuracy", "고정밀", "deep review", or equivalent - in ANY turn, even appended to a follow-up question and even after the plan already exists - set `review_required: true` in the draft: the dual high-accuracy review (native `momus` + the independent Oracle review) is now REQUIRED before handoff, and if the plan already exists you run it this same turn. The review runs under the bounded convergence contract in `full-workflow.md`: a 5-round cap (unlimited only on explicit user request), evidence-backed blocker eligibility, and approval-with-notes counting as approval. Answering the current question more carefully does NOT satisfy it. This does NOT choose CLEAR/UNCLEAR and does NOT suppress interview.
 
 After grounding, make ONE judgment, record `intent: clear|unclear` plus `review_required`, **ANNOUNCE both to the user in one line**, then load ONE intent reference (you ALSO read `references/full-workflow.md` for the shared mechanics - see below). The test keys on whether the desired **OUTCOME** is clear, NOT on request length. This verdict line and the opening announcement above are the two mandatory user-visible signals of a planning session - it tells the user whether they will be interviewed and whether high-accuracy review is already requested; never skip either.
 
@@ -43,7 +43,7 @@ After grounding, make ONE judgment, record `intent: clear|unclear` plus `review_
 
 - **OVERRIDE - explicit ask wins:** if the user explicitly asks to be questioned or interviewed ("ask me", "interview me", "why aren't you asking me" - in any language), route **CLEAR**, run the interview, and turn the adopt-default filter OFF: the user has claimed the forks, so every surviving one is ASKED, not defaulted. This beats the OUTCOME test below, even on a fuzzy brief.
 - **CLEAR** - the user knows the outcome; the only open items are preferences/tradeoffs the repo cannot answer (genuine owner-decisions). Read **`references/intent-clear.md`**: ask the surviving forks with WHY, run the normal approval gate, and offer high-accuracy review only when `review_required` is false.
-- **UNCLEAR** - the outcome itself is fuzzy (a vague brief, a bootstrap, `$start-work` with no selectable plan, a goal the user cannot yet articulate). Asking would offload your own job onto the user. Read **`references/intent-unclear.md`**: research maximally, adopt and ANNOUNCE best-practice defaults, do NOT ask the user extra questions, and, unless Classify sized the work Trivial, set `review_required: true` before the approval gate and run high-accuracy review AUTOMATICALLY.
+- **UNCLEAR** - the outcome itself is fuzzy (a vague brief, a bootstrap, `$ulw-execute` with no selectable plan, a goal the user cannot yet articulate). Asking would offload your own job onto the user. Read **`references/intent-unclear.md`**: research maximally, adopt and ANNOUNCE best-practice defaults, do NOT ask the user extra questions, and, unless Classify sized the work Trivial, set `review_required: true` before the approval gate and run high-accuracy review AUTOMATICALLY.
 - **ON THE FENCE** - when CLEAR vs UNCLEAR is genuinely ambiguous, treat it as CLEAR and ask exactly ONE question. A user wrongly silenced is worse than one extra question. The dominant failure to guard against is mis-routing a CLEAR request to UNCLEAR, which silently applies defaults and overrides forks the user wanted to own.
 
 WORKED: "add a 5/min-per-IP rate-limit to `/login`" = CLEAR. "make auth better" = UNCLEAR.
@@ -71,8 +71,7 @@ When producing the plan, encode every executable item as a column-zero Markdown
 - **Decision-complete is the north star.** The executor has NO interview context - spell out exact paths, "every X in Y", and an explicit Must-NOT-Have. Leave the implementer ZERO judgment calls.
 - **Full scope is the default.** Plan the ENTIRE request; "MVP", "v1", "phase 1", or any reduced subset is never an option you invent or ask about - it exists only if the user introduces it. Scope OUT / Must-NOT-Have entries are guardrails against unrequested additions, never reductions of the request.
 - **Explore before asking.** Discoverable facts (repo/system/docs truth) -> research and cite, never ask. Preferences/tradeoffs -> the only things you bring to the user. When unsure which, treat it as a user-decision.
-- **CodeGraph first when present.** Use `codegraph_explore` for repo how/where/what/flow questions before wider reads; if codegraph_* tools are absent, inactive/uninitialized, or cold-start unavailable, continue with Read/Grep/Glob/LSP and the ast-grep skill.
-- **Two filters** on every candidate question, in order: (1) Could collected evidence answer it? -> explore instead. (2) Could the user's stated intent plus a defensible default answer it? -> adopt the default, record it, do not ask - UNLESS it is an owner-decision, which always survives as a question even when a default exists: anything irreversible / destructive / safety-critical, or a cross-cutting product choice the user lives with (public config surface, distribution / packaging, external dependency or pinned SHA, data / schema shape). Default the reversible internals; surface the owner-decisions.
+- **Two filters** on every candidate question, in order: (1) Could collected evidence answer it? -> explore instead. (2) Could the user's stated intent plus a defensible default answer it? -> adopt the default, record it, do not ask - UNLESS it is an owner-decision, which always survives as a question even when a default exists: anything irreversible / destructive / safety-critical, or a cross-cutting product choice the user lives with (public config surface, distribution / packaging, external dependency or pinned SHA, data / schema shape, real budget / paid-service spend, expected scale or capacity target, target-audience / compliance limits). Extrinsic constraints (budget, mandated stack, scale, audience) leave no repo evidence, so exploration can never surface them - sweep those axes explicitly once per plan and classify each as explored, defaulted (ledger), or asked. Default the reversible internals; surface the owner-decisions.
 - **Explore to sufficiency, then STOP.** One research wave per open question; stop when the clearance check is answerable; never re-explore to double-check.
 - **Parallel-dispatch** independent research in ONE turn and keep working while it runs. Subagent outputs are CLAIMS until you independently verify them.
 - **Approval is not execution.** Approval authorizes writing the plan ONLY, never implementation. ONE request -> ONE plan, however large.