@1aboveio/skills 0.20.4 → 0.21.0

package/skills/content-creator/wechat-search/SKILL.md

@@ -0,0 +1,210 @@
+---
+name: wechat-search
+description: "Search WeChat 公众号 articles via Sogou Weixin, or read/summarize an article from a mp.weixin.qq.com URL (读/分析/总结) — suggest proactively when such a URL appears. Use for Chinese WeChat content, industry reports, competitor or trending topics. Requires Chrome Browser Relay. NOT general web search (web_search)."
+---
+
+# WeChat Article Search & Reader (搜狗微信搜索)
+
+Two modes:
+1. **Search mode** — search 公众号 articles via Sogou WeChat Search
+2. **Read mode** — read a full article from a `mp.weixin.qq.com` URL
+
+## Step 0: Ensure Chrome is Running (MANDATORY — run before anything else)
+
+**Every session must start here.** Do not skip this step.
+
+### Check if Chrome CDP is available
+
+```bash
+curl -s --max-time 2 http://localhost:18801/json/version
+```
+
+- **Returns JSON** → Chrome is ready, proceed to the workflow
+- **Fails / empty** → Chrome is not running, start it now (see below)
+
+### Start Chrome (Linux)
+
+Run the bundled script — it auto-detects the Chrome binary and waits until CDP is ready:
+
+```bash
+bash {baseDir}/scripts/start-chrome.sh
+```
+
+Expected output:
+```
+🚀 Starting Chrome...
+⏳ Waiting for CDP... ready!
+✅ Chrome CDP available at http://localhost:18801
+```
+
+If the script errors with "Chrome not found", install Chrome:
+```bash
+# Debian/Ubuntu
+sudo apt install google-chrome-stable
+# Fedora
+sudo dnf install google-chrome-stable
+# Snap
+sudo snap install chromium
+```
+
+### Start Chrome (macOS / Windows)
+
+**macOS:**
+```bash
+/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome \
+  --remote-debugging-port=18801 \
+  --user-data-dir=/tmp/chrome-cdp-profile \
+  --no-first-run &
+```
+
+**Windows:**
+```bat
+"C:\Program Files\Google\Chrome\Application\chrome.exe" ^
+  --remote-debugging-port=18801 ^
+  --user-data-dir=%TEMP%\chrome-cdp-profile ^
+  --no-first-run
+```
+
+> **Tip:** Reuse the same `--user-data-dir` across sessions to preserve WeChat login state.
+
+### Get tab ID (required for all browser calls)
+
+```
+browser action=tabs profile=chrome
+```
+
+Use the first `targetId` from the response in all subsequent `browser` calls.
+
+**Key parameters:**
+```
+profile=chrome    # CDP on localhost:18801 (real browser — avoids anti-bot)
+targetId=<id>     # from browser action=tabs
+```
+
+---
+
+## Mode 1: Read Article from URL
+
+Triggered when user provides a `mp.weixin.qq.com` URL.
+
+### Step 1: Try web_fetch first
+
+```
+web_fetch url="<mp.weixin.qq.com URL>"
+```
+
+If successful, extract and present:
+- **标题** — article title
+- **公众号** — account name (author)
+- **发布时间** — publish date
+- **正文** — full body text
+
+### Step 2: If web_fetch blocked (requires login), use browser
+
+```
+# Get attached tab ID
+browser action=tabs profile=chrome
+# → use the first targetId from the response
+
+browser action=navigate profile=chrome targetId=<id> targetUrl="<mp.weixin.qq.com URL>"
+# Wait ~2s for page to load
+browser action=snapshot profile=chrome targetId=<id>
+```
+
+Extract title, 公众号 name, date, and body text from the snapshot.
+
+### Output format (Read mode)
+
+```
+📄 **[标题]**
+公众号：[账号名] · [发布时间]
+
+[正文内容，完整呈现或摘要视长度而定]
+```
+
+---
+
+## Mode 2: Search Articles
+
+Triggered when user wants to search for WeChat content by keyword.
+
+### Step 1: Get tab ID
+
+Complete **Step 0** first. Then get the tab ID:
+
+```
+browser action=tabs profile=chrome
+```
+
+Use the first `targetId` for all subsequent browser calls.
+
+### Step 2: Navigate to Sogou WeChat Search
+
+```
+browser action=navigate profile=chrome targetId=<id> targetUrl="https://weixin.sogou.com/"
+```
+
+### Step 3: Snapshot to find search box
+
+```
+browser action=snapshot profile=chrome targetId=<id>
+```
+
+Look for the `textbox` element (search input).
+
+### Step 4: Type query and submit
+
+```
+browser action=act profile=chrome targetId=<id>
+request: { kind: "type", ref: "<search_box_ref>", text: "<query>", submit: true }
+```
+
+Wait ~2 seconds for results to load, then snapshot.
+
+### Step 5: Parse results
+
+```
+browser action=snapshot profile=chrome targetId=<id> compact=true
+```
+
+If no results (empty list or "未找到相关结果"), report "未找到与「<query>」相关的公众号文章" and suggest refining the keyword.
+
+Extract from snapshot:
+- **Title**: from `heading` elements (h3)
+- **Source** (公众号名): text after the article summary
+- **Date**: text next to source name
+- **Link**: the `/link?url=...` href (Sogou redirect URL)
+
+### Step 6: Pagination (optional)
+
+Click page links to get more results. Max 10 pages.
+
+### Search Types
+
+- **搜文章** (default): `https://weixin.sogou.com/` and search
+- **搜公众号**: navigate to `https://weixin.sogou.com/weixin?type=1&query=<keyword>`
+
+### Output format (Search mode)
+
+```
+1. **<Title>** — <Source> (<Date>)
+   <Link>
+2. **<Title>** — <Source> (<Date>)
+   <Link>
+...
+```
+
+---
+
+## Limitations
+
+- **Rate limiting**: Space out searches — no more than a few per minute
+- **Captcha**: If triggered, screenshot and ask the user to solve it manually
+- **Max 10 pages** per query (~100 results)
+- **Sogou redirect URLs expire** — extract real `mp.weixin.qq.com` URLs promptly if needed
+
+## Anti-Detection Tips
+
+- Use natural delays between actions
+- Don't paginate through all 10 pages rapidly
+- If captcha triggers, wait a few minutes before retrying

package/skills/content-creator/wechat-search/scripts/start-chrome.sh

@@ -0,0 +1,83 @@
+#!/usr/bin/env bash
+# start-chrome.sh — Launch Chrome with CDP enabled for weixin-search skill
+# Usage: ./start-chrome.sh [--port 18801] [--profile /tmp/chrome-cdp]
+
+set -euo pipefail
+
+CDP_PORT="${CDP_PORT:-18801}"
+USER_DATA_DIR="${USER_DATA_DIR:-/tmp/chrome-cdp-profile}"
+
+# Parse args
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --port) CDP_PORT="$2"; shift 2 ;;
+    --profile) USER_DATA_DIR="$2"; shift 2 ;;
+    *) echo "Unknown arg: $1"; exit 1 ;;
+  esac
+done
+
+# Find Chrome binary
+CHROME=""
+for candidate in \
+  google-chrome \
+  google-chrome-stable \
+  chromium \
+  chromium-browser \
+  /usr/bin/google-chrome \
+  /usr/bin/chromium \
+  /snap/bin/chromium; do
+  if command -v "$candidate" &>/dev/null || [ -x "$candidate" ]; then
+    CHROME="$candidate"
+    break
+  fi
+done
+
+if [ -z "$CHROME" ]; then
+  echo "❌ Chrome/Chromium not found. Install with:"
+  echo "   sudo apt install google-chrome-stable   # Debian/Ubuntu"
+  echo "   sudo dnf install google-chrome-stable   # Fedora"
+  echo "   sudo snap install chromium              # Snap"
+  exit 1
+fi
+
+# Check if CDP port already in use
+if curl -s --max-time 1 "http://localhost:${CDP_PORT}/json/version" &>/dev/null; then
+  echo "✅ Chrome already running on CDP port ${CDP_PORT}"
+  echo "   Tabs: http://localhost:${CDP_PORT}/json"
+  exit 0
+fi
+
+mkdir -p "$USER_DATA_DIR"
+
+echo "🚀 Starting Chrome..."
+echo "   Binary:   $CHROME"
+echo "   CDP port: $CDP_PORT"
+echo "   Profile:  $USER_DATA_DIR"
+
+"$CHROME" \
+  --remote-debugging-port="${CDP_PORT}" \
+  --user-data-dir="${USER_DATA_DIR}" \
+  --no-first-run \
+  --no-default-browser-check \
+  --disable-background-networking \
+  --disable-client-side-phishing-detection \
+  --disable-sync \
+  &
+
+# Wait for Chrome to become ready
+echo -n "⏳ Waiting for CDP..."
+for i in $(seq 1 15); do
+  if curl -s --max-time 1 "http://localhost:${CDP_PORT}/json/version" &>/dev/null; then
+    echo " ready!"
+    echo ""
+    echo "✅ Chrome CDP available at http://localhost:${CDP_PORT}"
+    echo "   Verify tabs: curl http://localhost:${CDP_PORT}/json"
+    exit 0
+  fi
+  echo -n "."
+  sleep 1
+done
+
+echo ""
+echo "⚠️  Chrome started but CDP not responding after 15s."
+echo "   Check manually: curl http://localhost:${CDP_PORT}/json/version"

package/skills/engineering/engineering-runtime/coherence/workflow.json

@@ -342,7 +342,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "eff6c7b5931bce5b2265a619bccddd371a89df6ce7bc2edc74f11d00060a71dd",
       "digestExcludes": []
@@ -355,7 +355,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "d97abeff7bdc5ae2e4be2f631d09dba46d4753ce1e67334793364329af59f066",
       "digestExcludes": []
@@ -368,7 +368,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "fe2ec3bae2bfec2dbef19db30ab0272391d7c1d9b6893341a76a9f339adefc57",
       "digestExcludes": []
@@ -381,7 +381,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "58b0556228a9271cf9e33727a05fc08a4fdfd29512ec0936b169f0a55d60e6ef",
       "digestExcludes": []
@@ -394,7 +394,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "d28dd4d3c4394df0cc7e6b48d0e5daa7d2277701ea459e39fba99acf1d8ff670",
       "digestExcludes": []
@@ -407,7 +407,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "56fb5843788c9073b0d5a65ffa72bd7d2fa80dc6a1ff06fb0e19fab513c51502",
       "digestExcludes": []
@@ -420,7 +420,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "f048fd00c69f2dc666fc7a3096cfeee6dfba73933bfb69602f3f0e26159798cc",
       "digestExcludes": []
@@ -433,7 +433,7 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
       "contentDigest": "d533684db4483cfc90be2aa7161980b6548057ef18a4b54f0cdbcd53c79dd46d",
       "digestExcludes": []
@@ -446,15 +446,15 @@
         "id": "first-party",
         "type": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4"
+        "version": "0.21.0"
       },
-      "contentDigest": "86a91b567eabe6143a4ecac3cbec7271da252978202f292e287bb3a8da95a797",
+      "contentDigest": "e4a907bd6c3eec9fbfb1c03e785c0e58b81fbae6e2d739469ea24be2be748e45",
       "digestExcludes": [
         "coherence/workflow.json"
       ]
     }
   ],
-  "releaseIdentity": "ffa735d74d402c79fbd8cbca332b415422884ee88f914412d4aa88c2d1353cb9",
+  "releaseIdentity": "d2ddb6d571f5b326aca54fbb7c11b5417f4ad23d09842ee629f0181138cd6d11",
   "lifecycleAuthority": "native-skills-cli",
   "repairRecipe": {
     "id": "engineering-workflow-dependency-first",
@@ -512,7 +512,7 @@
         "sourceId": "first-party",
         "sourceType": "first-party",
         "package": "@1aboveio/skills",
-        "version": "0.20.4",
+        "version": "0.21.0",
         "installPath": null,
         "members": [
           "harness-runtime",
@@ -528,7 +528,7 @@
         "commands": [
           {
             "transport": "npm",
-            "command": "npx @1aboveio/skills@0.20.4 install --group engineering-workflow --yes"
+            "command": "npx @1aboveio/skills@0.21.0 install --group engineering-workflow --yes"
           }
         ],
         "onFailure": {

package/skills/engineering/engineering-runtime/scripts/workflow-policy.mjs

@@ -86,7 +86,7 @@ export const WORKFLOW_TRUSTED_SOURCES = deepFreeze({
     id: 'first-party',
     type: 'first-party',
     package: '@1aboveio/skills',
-    version: '0.20.4',
+    version: '0.21.0',
   },
   matt: {
     id: 'matt-pocock',