MarketIntelRadarSkill

---
name: MarketIntelRadarSkill
description: >
  Compile multi-source market intelligence into weekly digests with priority scoring and response suggestions.
version: 0.9.0
visibility: organization
tags:
  - marketing
  - intelligence
  - research
entrypoint:
  command: compile_market_digest
  label: Build weekly intelligence brief
---

## When Claude should load this skill
- Trigger each Monday 03:00 using `cron: 0 3 * * MON` or run manually after key announcements.
- Ideal when stakeholders request consolidated news, sentiment signals, and recommended follow-up actions.

## Core workflow
1. Review `instructions/market-radar.md` to confirm data sources, weighting logic, and alert thresholds.
2. Execute `python scripts/market_radar.py` to emit `MARKET_RADAR` with prioritized headlines and mention trends.
3. Populate `templates/weekly-report-outline.md`, tailoring call-to-actions for marketing, sales, and comms teams.

## Linked resources
- `data/news_feed.csv` — Curated industry headlines with impact and relevance scores.
- `data/social_mentions.csv` — Social listening exports grouped by tag, volume, and sentiment.
- `instructions/market-radar.md` — Guidance on classification labels, scoring heuristics, and escalation rules.
- `templates/weekly-report-outline.md` — Executive-ready briefing structure for distribution.
- `dashboards/share-of-voice.csv` — Optional KPI snapshot for long-term trend comparison.

import pandas as pd

news = pd.read_csv("data/news_feed.csv")
mentions = pd.read_csv("data/social_mentions.csv")

news["score"] = news["impact"] * news["relevance"]
priority_news = news.nlargest(5, "score")

mentions_grouped = (
    mentions.groupby("tag")["count"].sum().reset_index().sort_values("count", ascending=False)
)

digest = {
    "headline_highlights": priority_news[["title", "source", "score"]].to_dict("records"),
    "top_mentions": mentions_grouped.head(5).to_dict("records"),
}

print("MARKET_RADAR=", digest)