Pangolinfo AI SERP Skill

Search Google and retrieve AI Overviews, organic search results, and screenshots via Pangolin APIs.

When to Use This Skill

Activate this skill when the user's intent matches any of these patterns:

• "Search Google for ...", "Google ...", "Look up ..."
• "What does Google say about ...", "Search the web for ..."
• "Get AI overview for ...", "AI search for ..."
• "Get search results for ...", "SERP for ..."
• "Take a screenshot of Google results for ..."
• "Follow up on the search with ..."
• Chinese equivalents: "搜索一下...", "谷歌搜索...", "帮我查一下...", "用Google搜...", "搜一下...", "AI搜索..."

Do not use this skill for Amazon product searches, price lookups, or review scraping -- those require a different skill.

Prerequisites

Runtime

• Python 3.8+ (uses only the standard library -- no pip install needed)

Pangolin Account

Register at pangolinfo.com to obtain credentials.

Environment Variables

Set one of the following:

API key resolution order: PANGOLIN_API_KEY env var > cached ~/.pangolin_api_key > fresh login.

macOS SSL Certificate Fix

On macOS, Python may fail with CERTIFICATE_VERIFY_FAILED because it ships without root certificates by default.

Symptoms: The script outputs an error with code SSL_CERT.

Solutions (pick one):

1. Run the certificate installer that ships with Python:

   /Applications/Python\ 3.x/Install\ Certificates.command
   

   (Replace 3.x with your Python version, e.g. 3.11)

2. Set the SSL_CERT_FILE environment variable:

   pip3 install certifi
   export SSL_CERT_FILE=$(python3 -c "import certifi; print(certifi.where())")
   

First-Time Setup Guide

When a user tries to use this skill and authentication fails (error code MISSING_ENV), do not just repeat the error hint. Instead, walk the user through the full setup process interactively:

Step 1: Explain what's needed

Tell the user (in their language):

To use this skill, you need a Pangolin API account. Pangolin provides Google search and AI Overview data through its APIs.

使用本技能需要 Pangolin API 账号。Pangolin 提供 Google 搜索和 AI 概览数据的 API 服务。

Step 2: Guide registration

1. Go to pangolinfo.com and create an account

2. After login, find your API Key in the dashboard

3. 访问 pangolinfo.com 注册账号

4. 登录后在控制台找到你的 API Key

Step 3: Collect credentials and authenticate automatically

When the user provides their credentials, you (the AI agent) should configure them securely. The script will automatically cache the API key at ~/.pangolin_api_key for all future calls.

If user provides an API key (recommended): Write it directly to the cache file — avoids shell history entirely:

echo "<api_key>" > ~/.pangolin_api_key
chmod 600 ~/.pangolin_api_key 2>/dev/null
python3 scripts/pangolin.py --auth-only

If user provides email + password: Set env vars in the session and clean up after auth:

export PANGOLIN_EMAIL="user@example.com"
export PANGOLIN_PASSWORD="their-password"
python3 scripts/pangolin.py --auth-only
unset PANGOLIN_EMAIL PANGOLIN_PASSWORD

This avoids passwords appearing in shell history (unlike inline VAR=x command syntax) and cleans up credentials after the API key is cached.

Both methods cache the API key automatically. After this one-time setup, no environment variables are needed — all future calls will use the cached API key at ~/.pangolin_api_key.

Step 4: Confirm and proceed

After auth returns "success": true:

1. Tell the user: "认证成功！API Key 已自动缓存，后续使用无需再次输入。" / "Authentication successful! API key cached — no need to enter credentials again."
2. Immediately retry their original request.

Important

• The user only needs to provide credentials ONCE — the script caches the API key permanently at ~/.pangolin_api_key
• Do not ask the user to manually edit .bashrc or .zshrc — the script handles persistence automatically
• If the user doesn't have an account yet, explain Pangolin's credit system (2 credits per AI Mode search, 0.5 credits per SERP search) and direct them to pangolinfo.com
• If auth succeeds but credits are exhausted (error code 2001), tell the user to top up at pangolinfo.com
• API key is permanent and does not expire unless the account is deactivated

Script Execution

The main script is located at scripts/pangolin.py relative to this skill directory.

Path resolution: When invoking the script, resolve the absolute path from this skill's directory. Example:

SKILL_DIR="$(cd "$(dirname "$0")" && pwd)"
python3 "$SKILL_DIR/scripts/pangolin.py" --q "your query"

Or invoke directly:

python3 scripts/pangolin.py --q "your query"

Intent-to-Command Mapping

AI Mode Search (default)

User wants AI-generated answers with references. This is the default mode.

python3 scripts/pangolin.py --q "what is quantum computing"

Standard SERP

User wants traditional Google search results (organic links + optional AI overview).

python3 scripts/pangolin.py --q "best databases 2025" --mode serp

SERP with Screenshot

User wants to see the actual Google results page.

python3 scripts/pangolin.py --q "best databases 2025" --mode serp --screenshot

SERP with Region

User wants SERP results from a specific geographic region.

python3 scripts/pangolin.py --q "best databases 2025" --mode serp --region us

Multi-Turn Dialogue

User wants to ask follow-up questions in an AI Mode conversation.

python3 scripts/pangolin.py --q "kubernetes" --follow-up "how to deploy" --follow-up "monitoring tools"

Auth Check

Verify credentials are working without consuming credits.

python3 scripts/pangolin.py --auth-only

Smart Defaults

When the user simply says "search for X" without specifying a mode, use the default ai-mode.

Output Format

The script outputs JSON to stdout on success and structured error JSON to stderr on failure.

Success Example (AI Mode)

{
  "success": true,
  "task_id": "1768988520324-766a695d93b57aad",
  "results_num": 1,
  "ai_overview_count": 1,
  "ai_overview": [
    {
      "content": ["Quantum computing uses quantum bits (qubits)..."],
      "references": [
        {
          "title": "Quantum Computing - Wikipedia",
          "url": "https://en.wikipedia.org/wiki/Quantum_computing",
          "domain": "Wikipedia"
        }
      ]
    }
  ],
  "screenshot": "https://image.datasea.network/screenshots/..."
}

Success Example (SERP)

{
  "success": true,
  "task_id": "1768988520324-abcdef123456",
  "results_num": 3,
  "ai_overview_count": 1,
  "ai_overview": [
    {
      "content": ["Java works by compiling source code..."],
      "references": [
        {"title": "How Java Works", "url": "https://docs.oracle.com/...", "domain": "Oracle"}
      ]
    }
  ],
  "organic_results": [
    {
      "title": "Java Tutorial for Beginners",
      "url": "https://example.com/java-tutorial",
      "text": "Learn how Java works from the ground up..."
    }
  ]
}

Error Example

{
  "success": false,
  "error": {
    "code": "MISSING_ENV",
    "message": "No authentication credentials found.",
    "hint": "Set PANGOLIN_API_KEY, or both PANGOLIN_EMAIL and PANGOLIN_PASSWORD environment variables."
  }
}

Response Presentation

When presenting results to the user:

1. Use natural language -- never dump raw JSON to the user.
2. Match the user's language -- if the user writes in Chinese, respond in Chinese; if English, respond in English.
3. Summarize the AI overview first (if present), then list relevant organic results with titles and URLs.
4. Include source URLs so the user can verify or explore further.
5. Mention the screenshot URL if one was captured, so the user can view the rendered page.
6. If no AI overview is present, note that Google did not generate one for this query and present organic results instead.
7. On error, explain the issue in plain language and suggest corrective action based on the hint field.

All CLI Options

Output Schema

See references/output-schema.md for the complete JSON output schema documentation.

Cost

• AI Mode: 2 credits per request
• SERP: 0.5 credits per request

Credits are only consumed on successful requests (API code 0). Auth checks (--auth-only) do not consume credits.

Exit Codes

Troubleshooting

Error Communication Table

Pangolin API Error Codes

Important Notes for AI Agents

1. Always run --auth-only first if you are unsure whether credentials are configured, before spending credits on a real query.
2. Default to ai-mode unless the user explicitly asks for standard/traditional search results.
3. Never expose raw JSON to the user. Parse the output and present it in natural language.
4. Respect the user's language. Respond in the same language the user is writing in.
5. Keep follow-ups to 5 or fewer for optimal response time. Warn the user if they request more.
6. Handle errors gracefully. If the script exits with a non-zero code, read stderr for the structured error and present the hint to the user.
7. Do not log or echo API keys, passwords, or cookies in any output, logs, or error messages.
8. Credit awareness: AI Mode costs 2 credits per search; SERP costs 0.5 credits. If a user requests many searches in a row, mention the credit cost.
9. Screenshot is optional. Only pass --screenshot when the user explicitly wants to see the rendered page or when visual context is needed.
10. Multi-turn is ai-mode only. Do not attempt --follow-up with --mode serp.