yang0/skill_grok/grok-query-scrape

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Submit prompts to https://grok.com using an authenticated cookie file and capture the generated answer verbatim. Use when the task requires raw prompt passthrough, raw Grok output capture, and reproducible artifacts without JSON parsing.

Install

npx vskill@latest install yang0/skill_grok/grok-query-scrape

SKILL.md

Grok Query Scrape

Use this skill to run a repeatable Grok automation workflow:

  • Load Netscape-format cookies (for logged-in session reuse)
  • Open Grok in a browser session
  • Submit the prompt exactly as provided
  • Wait for answer generation to stabilize
  • Return and save raw output only (no parsing)

Language guides:

Workflow

  1. Install runtime dependencies when missing:
powershell
pip install patchright
  1. Prepare a cookie file in Netscape format (example: H:\cookies\grok.txt).
  1. Run the script:
powershell
python E:\projectHome\skill_grok\skill\grok-query-scrape\scripts\query_grok.py `
--cookie-file H:\cookies\grok.txt `
--prompt "Find top 10 AI productivity tweets from the last 24 hours." `
--output-dir E:\ `
--basename grok_ai_top10 `
--wait-seconds 240
  1. Check outputs:
  • <stdout>: Grok response text (verbatim capture from main)
  • <basename>_raw.txt: same raw Grok response text saved to file
  • <basename>_meta.json: run metadata and paths

Script

Use query_grok.py for execution.

Key flags:

  • --cookie-file: Netscape cookie txt path (required)
  • --prompt: inline prompt
  • --prompt-file: prompt text file (alternative to --prompt)
  • --output-dir: output folder (default .)
  • --basename: output filename prefix (default grok_result)
  • --wait-seconds: max wait for response stabilization (default 180)
  • --headless: run browser headless
  • --new-chat: send Ctrl+J before query to reduce contamination (enabled by default)

Prompting Guidance

This skill forwards the prompt as-is, so put all constraints directly into your prompt text.

Failure Handling

If the output format is not what you need:

  1. Keep _raw.txt as ground truth.
  2. Tighten the prompt and retry.
  3. Keep --new-chat enabled for isolation.
  4. Increase --wait-seconds when generation is long.

If login fails:

  1. Refresh cookie file from a valid browser session.
  2. Keep domain/path/secure flags intact in Netscape format.
  3. Retry with --headless disabled to inspect behavior.
scripts/
  • query_grok.py: submit prompt, wait for answer, return/save raw/meta only.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/yang0/skill_grok/grok-query-scrape/badge)](https://verified-skill.com/skills/yang0/skill_grok/grok-query-scrape)

Something wrong with this skill or its page? Tell us.