trend-radar
Skill4 starsChanged 4 months agoMITView on GitHub
Scheduled Monday-morning trend radar — scans the user's chosen topic for 5 fresh signals, each triangulated across 2+ independent sources, with a so-what and suggested action per signal. Use when the user asks to "run my trend radar", "what's new in <my space> this week", "set up weekly market monitoring", or wants recurring awareness of a topic. Don't use for a one-off deep dive on a single competitor — that's competitor-teardown — or for sizing the market — that's market-sizing-tam.
4 other projects use a version of this.
Adopted by
Most recently taken up by
- neubauerwilli/gta6-skilltheir own version, first seen Nov 15, 2026
- Cheukcoding/claude-skillstheir own version, first seen Nov 1, 2026
- easyaitech/etsy-skillstheir own version, first seen Oct 16, 2026
- scrollmark/social-skillstheir own version, first seen Oct 14, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- scrollmark/social-skills★ 19same name
- neubauerwilli/gta6-skill★ 10same name
- easyaitech/etsy-skills★ 5same name
- Cheukcoding/claude-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository74 other files it gives its agents.
- agent-eval-harnessSkill
- context-budget-managerSkill
- incident-rollbackSkill
- memory-hygieneSkill
- model-cost-optimizerSkill
- prompt-debuggerSkill
- skill-author-coachSkill
- third-party-skill-vetterSkill
- api-endpoint-testerSkill
- bug-triage-protocolSkill
- changelog-release-notesSkill
- docs-readme-writerSkill
- env-secrets-hygieneSkill
- github-pr-reviewSkill
- git-ship-flowSkill
- lovable-brief-writerSkill
- supabase-migration-writerSkill
- vercel-deploy-checkSkill
- cold-outreach-sequencerSkill
- competitor-pricing-scanSkill
- contract-redline-scanSkill
- customer-support-inboxSkill
- invoice-runnerSkill
- kpi-dashboard-weeklySkill
- lead-research-dossierSkill
- payment-reconciliationSkill
- pipeline-trackerSkill
- proposal-builderSkill
- stripe-payment-linkSkill
- testimonial-harvesterSkill
- weekly-revenue-reportSkill
- brand-voice-guardianSkill
- case-study-builderSkill
- content-repurposerSkill
- deep-research-briefSkill
- email-nurture-sequenceSkill
- landing-page-copywriterSkill
- linkedin-content-packSkill
- newsletter-assemblerSkill
- seo-blog-writerSkill
- social-listening-reportSkill
- x-thread-builderSkill
- youtube-script-writerSkill
- budget-guardrailsSkill
- ceo-operating-doctrineSkill
- delegation-protocolSkill
- editor-personaSkill
- heartbeat-standupSkill
- hiring-briefSkill
- ops-agent-personaSkill
- researcher-personaSkill
- sales-agent-personaSkill
- support-agent-personaSkill
- weekly-retroSkill
- writer-personaSkill
- calendar-conciergeSkill
- decision-journalSkill
- email-triageSkill
- file-organizerSkill
- habit-trackerSkill
- The first 60 shown.
The file
Files
Trend Radar
A blueprint skill: every Monday at 08:00 it scans the user's chosen topic and delivers exactly 5 signals — each triangulated across at least 2 independent sources, dated, and finished with a so-what and one suggested action. The standard: a signal with one source is a rumor and doesn't ship; a signal without a so-what is noise wearing a suit. Read-only web work, safe to run unattended.
When to Use
- The Monday scheduled run (automatic via the blueprint above).
- Ad-hoc: "what happened in
<topic>this week?", "run the radar now". - Setting up: user names a topic they want watched weekly.
- Not for: deep single-company analysis (
competitor-teardown); market size math (market-sizing-tam); auditing one page (seo-content-audit).
Quick Reference
| Action | Command / Call |
| Fresh news | web search: <topic> ("launched" OR "announced" OR "raised" OR "shut down"), filter to last 7–14 days |
| Community pulse | search <topic> site:reddit.com and site:news.ycombinator.com — sort/filter recent |
| Primary sources | web_extract the actual announcement/changelog/paper, not the article about it |
| Triangulate | second independent source per signal — different outlet/author, not a syndication |
| Date check | every source needs a publish date within the window; undated → doesn't count |
| Memory | compare against last week's radar (from notes/memory) to catch signal evolution |
Procedure
- Precheck (first run only) — confirm with the user: the topic (specific enough to search — "AI agents for real-estate teams", not "AI"), 3–6 seed sources they trust, and what decisions the radar feeds (build/marketing/investing). Store in memory. On scheduled runs, load these and go; if topic config is missing, deliver a setup request instead of a generic scan.
- Sweep — run 4–6 searches: the news pattern, the community pattern, the seed sources' recent output, and one wildcard angle ("
<topic>problems", "<topic>alternatives"). Collect candidate items with URL + publish date. Window: 7 days (stretch to 14 if thin, and say so). - Shortlist — keep candidates that are new (not in last week's radar), dated inside the window, and relevant to the user's stated decisions. Drop press-release fluff with no substance.
- Triangulate — for each shortlisted item, find a second independent source (different outlet or the primary source itself). Syndicated copies of the same wire story count as one source. Can't find a second? The item ships only if flagged
single-source — unconfirmed, and at most one such item per radar. - Rank to 5 — score by relevance to the user's decisions × strength of evidence. Exactly 5 signals: fewer available means saying "quiet week — N real signals" rather than padding; more means the rest go in a one-line "also noticed" footer.
- So-what and action — per signal, one sentence on why it matters to this user, and one concrete suggested action (test, post, outreach, watch-item) with a confidence label. Generic advice ("keep an eye on this") is banned.
- Deliver — fill the template, save this week's signal list to memory for next week's comparison, and send through the run's normal delivery channel. On any failure (web down, no config), still deliver a short note saying why — the radar is never silently absent.
Output Template
# Trend Radar — <topic> — week of <date>
Window: <dates> | Searches run: N | Candidates: N → shipped: 5
## Signals
1. **<headline in plain words>** (<date>)
Sources: <url1> + <url2> [or: single-source — unconfirmed]
So what: <why this matters to YOUR decisions>
Suggested action: <one concrete step> (confidence: high/med/low)
2–5. ...
## Movement since last week
- <signal from a past radar that advanced/died — or "first run, no baseline">
## Also noticed
- <one-liners that didn't make the cut>Pitfalls
- Old news dressed as new — search surfaces a 2024 article and it reads like this week's signal. Recovery: hard date-check every source; undated pages are excluded; the radar header states the window and every signal carries its date.
- One press release echoing across ten outlets — looks triangulated, is actually one source syndicated. Recovery: triangulation requires independent authorship — check bylines and quoted material; syndications collapse to one.
- Topic drift — after a few weeks the radar quietly becomes generic industry news. Recovery: re-read the stored decision context each run; every so-what must reference the user's decisions, and a quarterly line asks the user if the topic config still fits.
- Padding a quiet week — 5 signals demanded, 2 exist, 3 get invented from thin air. Recovery: quiet weeks report honestly ("2 real signals this week"); credibility of the radar is worth more than the count.
- Scheduled run fails silently — web toolset hiccup means no Monday radar and nobody notices for a month. Recovery: on any failure, deliver a one-line failure note with the reason and the fix, so absence is always visible.
Verification
- Exactly 5 signals or an explicit "quiet week" count — never padded
- Every signal: 2+ independent sources with URLs and dates inside the window (max one flagged single-source)
- Every so-what references the user's stored decision context
- Every action is concrete and confidence-labeled
- This week's signals saved to memory; last week's compared
- Failure note delivered if the run couldn't complete
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

