radar-merge-briefing
Skill4 starsChanged 5 months agoMITView on GitHub
Merges all competitor dossiers and the change set into one briefing. Writes three local artifacts: the briefing markdown, a combined raw dossier JSON, and a dashboard data JSON. Use after radar-diff-changes and before radar-publish-notion and radar-render-dashboard. Every claim in the briefing carries a source URL.
In this repository8 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- radar-diff-changesSkill
- radar-onboardSkill
- radar-plan-targetsSkill
- radar-provision-notionSkill
- radar-publish-councilSkill
- radar-publish-notionSkill
- radar-render-dashboardSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 27
No claim without a source (no-fabrication.md).
- Distinctive: line 29
No em or en dashes (no-em-dashes.md).
- Very specific: line 38
Update radar.config.json last_run with the new paths and date, and roll the previous run's dossier dir so the next diff has a baseline.
- Distinctive: line 87
All competitors blocked: still produce a briefing that honestly says data could not be gathered and suggests checking the BrightData …
- Distinctive: line 89
Keep the dashboard data JSON in sync with the markdown so all three surfaces agree.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
radar-merge-briefing
Automation level: AUTONOMOUS.
Goal: produce the single source of truth for this run on local disk, with a source URL on every claim and no fabricated or stale-as-fresh values.
Steps
- Inputs from the caller: the run date, the run dossier directory, and the change set from
radar-diff-changes. Readradar.config.jsonfor business name, keywords, and freshness windows. - Load every dossier in the run directory.
- Render the briefing markdown from the template at
${CLAUDE_PLUGIN_ROOT}/assets/briefing-template.md. Fill every placeholder. For each competitor field, write the value, its source link, the fetched date, and the status. Where a field isnot_foundorblocked, say so plainly rather than leaving a gap or inventing a value. Build the "What changed" and "Focus list" sections from the ranked change set. - Enforce the rules as you write:
- No claim without a source (no-fabrication.md).
- Label freshness per section (freshness.md).
- No em or en dashes (no-em-dashes.md); normalize any in quotes.
- Write three files:
- Briefing markdown:
outputs/briefings/<YYYY-MM-DD>-briefing.md - Combined raw dossier JSON:
outputs/dossiers/<YYYY-MM-DD>/combined.json(all dossiers plus the change set, the machine-readable record). - Dashboard data JSON:
outputs/dashboard/<YYYY-MM-DD>-data.json, shaped for the dashboard template (see the shape below).war_roomis null until/radar-councilfills it.
- Briefing markdown:
- Update
radar.config.jsonlast_runwith the new paths and date, and roll the previous run's dossier dir so the next diff has a baseline.
Dashboard data JSON shape
The dashboard at ${CLAUDE_PLUGIN_ROOT}/assets/dashboard-template.html reads this exact shape. The
template computes missing KPIs itself, but populate kpis so the numbers match the briefing. SERP
rank values MUST be numbers (not strings) so the leaderboard renders. Sentiment values are one of
positive, mixed, negative, unknown. Each change carries a type
(RISING, FALLING, PRICE, NEWS, NEW, STEADY) and the competitor name.
{
"business_name": "Acme Co",
"run_date": "2026-06-16",
"keywords": ["primary keyword", "secondary keyword"],
"kpis": { "competitors_tracked": 6, "rising_threats": 2, "price_moves": 1,
"new_signals": 3, "cited_pct": 100 },
"changes": [
{ "type": "RISING", "competitor": "Competitor One",
"text": "moved to SERP #2 for \"primary keyword\" (was #5).", "source_url": "..." }
],
"competitors": [
{
"name": "Competitor One", "domain": "competitor-one.com",
"threat_level": "", "change": "Rising",
"serp_rank": { "value": 2, "keyword": "primary keyword", "source_url": "...", "fetched_at": "...", "status": "ok" },
"pricing": { "value": "Pro $49/user/mo", "source_url": "...", "fetched_at": "...", "status": "ok" },
"positioning": { "value": "...", "source_url": "...", "fetched_at": "...", "status": "ok" },
"news": { "value": "...", "source_url": "...", "fetched_at": "...", "status": "ok" },
"sentiment": { "value": "mixed", "source_url": "...", "fetched_at": "...", "status": "ok" },
"funding": { "value": null, "source_url": null, "fetched_at": "...", "status": "not_found" },
"hiring": { "value": null, "source_url": null, "fetched_at": "...", "status": "not_found" }
}
],
"war_room": null
}threat_level stays empty at gather time (low/medium/high is assigned by the council) and the
dashboard simply omits the threat chip and filter until it is set. change comes from
radar-diff-changes (New, Rising, Falling, Steady).
- Return the three local paths to the caller.
Edge cases
- A dossier file is malformed: skip it, note the competitor as "dossier unreadable this run," and continue. Do not crash the whole briefing.
- All competitors blocked: still produce a briefing that honestly says data could not be gathered and suggests checking the BrightData connection.
- Keep the dashboard data JSON in sync with the markdown so all three surfaces agree.
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.

