AI AGENT SKILLS

Phy Competitor Watch

一个面向 Content 场景的 Agent 技能。原始说明:Competitor social media content strategy analyzer. Feed it a JSON file of any competitor's posts and get a full strategy breakdown — topic distribution, cont...

SKILL.md

SKILL.md


name: Competitor Watch
description: Competitor social media content strategy analyzer. Feed it a JSON file of any competitor's posts and get a full strategy breakdown — topic distribution, content style analysis (listicle/story/contrarian/how-to/data-driven/thread), posting schedule, top-performing content patterns, and exploitable gaps (topics and styles they DON'T cover). Works with any platform's exported data. No API keys needed. Zero external dependencies.
license: Apache-2.0
homepage: https://canlah.ai
metadata:
author: Canlah AI
version: "1.0.3"
tags:

  • social-media
  • competitive-intelligence
  • content-strategy
  • analytics
  • marketing

phy-competitor-watch — Competitor Content Strategy Analyzer

Feed it a competitor's posts, get their entire content strategy decoded.

python3 ~/.claude/skills/phy-competitor-watch/scripts/competitor_watch.py \
  --file competitor_posts.json --name "Dan Koe"

What It Analyzes

  • Topic distribution — what they write about most (AI/ML, DevTools, StartupGrowth, etc.)
  • Content styles — listicle, story, contrarian, how-to, data-driven, question-led, announcement, thread
  • Posting schedule — which days they post, frequency per week
  • Top performers — their best posts by engagement + what made them work
  • Gaps you can exploit — topics and styles they DON'T cover

Input Format

[
  {"text": "Post content...", "platform": "linkedin", "date": "2026-03-01", "engagement": 5.2},
  {"text": "Another post...", "date": "2026-02-28", "engagement": 12.1}
]

Collecting Competitor Data

Use phy-twitter-scrape, Reddit API exports, or manual collection. The tool works with any JSON array of posts.


Author

Canlah AI — Run performance marketing without breaking your brand.