Auto-fetch competitor moves daily, structure them with Kimi, store in Feishu-n8n orchestrates the whole thing, humans only do 15 minutes of review. This is the competitor-tracking workflow AI Cake uses itself-import it, swap the API key, and run.
Workflow Chain
Scheduled trigger -> fetch competitor notes -> Kimi structured rewrite -> Feishu storage -> notify.
Download Template
After importing into n8n, configure per the steps below.
Steps
- Import: n8n -> Workflows -> Import from File, pick competitor-tracker.json
- Cron node: Set to fire daily at 9:00 (or your preferred frequency)
- HTTP Request (fetch): Fill your compliant data source URL-bring your own (don't scrape copyrighted content)
- Kimi API node: Fill the Kimi API key (replace placeholder sk-xxx with yours), pick model moonshot-v1-8k
- Feishu node: Create a Feishu Bitable, fill App ID / Secret + table token
- Try run: Manually trigger once, check if the Feishu table received data
Companion Prompt (Kimi Structuring)
You are a senior topic editor. Please extract from this note:
1. Title pattern (number / pain point / contrast)
2. Body structure (opening - middle - ending)
3. Reusable topic angle
Output JSON only, fields: title_pattern, structure, anglePitfalls
- Kimi occasionally returns non-JSON: Stress "output JSON only" at the end of the prompt; add a JSON-parse retry branch in n8n
- Scrape rate too high gets rate-limited: Once per day per competitor is enough-don't get greedy
- Feishu fields must be created first: n8n won't auto-create fields; a missing field fails silently
The template is the skeleton; the data source and API key are your own. Once it runs, you can add nodes: store price trends in a database, push anomaly alerts to WeCom.