Help us improve
Share bugs, ideas, or general feedback.
From claude-plugin
Generate a comprehensive, visual account research dashboard for any company using the fast reo_get_account_research composite endpoint. Use this skill whenever the user asks to "research an account", "pull up account intel", "get me a brief on [company]", "show me what's going on at [company]", "run account research", or anything similar. Also trigger this skill when the user mentions a company name or domain and wants sales context, buyer intel, hiring signals, engagement activity, or contact information. Even vague prompts like "what do we know about [company]?" or "prep me for my call with [company]" should trigger this skill. Output is always a dark-mode HTML dashboard rendered inline.
npx claudepluginhub reodotdev/claude-plugin --plugin reo-mcp-toolsHow this skill is triggered — by the user, by Claude, or both
Slash command
/claude-plugin:account-research-v2The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Generate a full-page, dark-mode HTML account intelligence dashboard. All reo data comes from a
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Create and present web-based slidedecks using Slidev with Markdown, Vue components, code highlighting, animations, interactive demos, LaTeX, and Mermaid for technical presentations, conference talks, code walkthroughs, and teaching materials.
Share bugs, ideas, or general feedback.
Generate a full-page, dark-mode HTML account intelligence dashboard. All reo data comes from a
single composite call — reo_get_account_research — and web searches supplement it with external
context. The skill's job is to compose everything together and render the dashboard.
Claude's job:
reo_get_account_research once to get all reo data (firmographic, scores, activities, developers, prospects, hiring)If the user gives a company name but not a domain, infer the most likely domain (e.g. "Vercel" → "vercel.com"). If ambiguous, ask before proceeding.
Call reo_get_account_research. Use a 90-day signal window, max 5 contacts per section, and
standard high-intent page labels:
reo_get_account_research(
domain = "<company domain>",
signal_lookback_days = 90,
max_contacts = 5,
high_intent_pages = ["Website: Pricing", "Form: Cloud PAYG", "Form: Self-hosted PAYG", "Form: Contact Sales", "Website: Case studies"],
prospect_seniority_filter = ["VP", "Director", "Head", "C-Level"]
)
This is the only reo call. Everything in the dashboard is driven by this response:
| Response Section | What It Contains |
|---|---|
meta.org_id | UUID for the account |
firmographic | Company name, industry, HQ, tech headcount, tech stack, customer_fit_score (ICP tier), developer_stage, developer_activity_score, reo_account_link |
engagement | Intent level, event counts, high-intent hits, activity type breakdown |
top_activities | Recent activity feed — use to render the Activity Summary section |
developers | Up to 5 deanonymized developers with email, activity summary, notable pages |
prospects | Up to 5 LinkedIn prospects with title, seniority, LinkedIn URL |
hiring | Total open roles, breakdown by tech function, date range |
If any section is empty, display "— Not available" in the dashboard rather than skipping the section.
Run these three searches in parallel — they supplement the reo data with external context:
Search for tools/vendors the company uses in your category. Look for job descriptions, LinkedIn posts, tech blog posts. For each competitor found: name, status (Active/Primary, Partial, Churned, Complementary), 2-sentence description. Include objection prep if a primary competitor is found.
Search for news from the last 30 days: funding, exec hires, product launches, outages. For each event: icon, title, 1-sentence sales implication, date. Omit section if nothing found.
Use the contacts found in the reo response to identify buyer personas. For each, state their likely KPIs based on role + industry. Brief web search if needed for industry-specific KPIs.
Use firmographic.customer_fit_score (e.g. "STRONG", "MODERATE", "WEAK") directly. No custom
scoring — the composite endpoint is the source of truth. Display the tier prominently and list
the supporting signals from the firmographic/hiring data (industry, tech headcount, hiring
shape, dev stage) alongside it as evidence.
These are where Claude adds real value beyond raw data. Be specific and opinionated — tie observations to actual people, actions, and dates from the data. No generic filler.
2–4 sentences on what the activity pattern means. Consider: velocity (accelerating/stalling?), breadth (one person or many?), depth (deep in docs or just browsing?), intent signals (pricing/demo hits?). Connect to a likely sales scenario.
Synthesise what the collective developer behaviour signals. Examples of the kind of specificity expected:
3–5 concrete, prioritised actions. Each includes:
Order by urgency. Highest-signal, lowest-effort first. If the account is cold, suggest prospecting actions based on hiring signals and contacts.
Generate a full dark-mode HTML dashboard. Self-contained (all CSS inline, no external deps).
Color palette:
#1a1a1a (page), #242424 (card), #2d2d2d (nested card)#e8e8e8; secondary: #9a9a9a; label: #6b6b6b (uppercase, 11px, 0.08em tracking)#4ade80; blue: #60a5fa; orange: #f59e0b; red: #f87171; purple: #a78bfa#3a3a3a; ICP fill: #4ade80; Activity fill: #60a5fa#f87171, Partial=#f59e0b, Churned=#4ade80, Complementary=#9a9a9a#14532d, green text #4ade80#14532d bg, #4ade80 text#422006 bg, #f59e0b text#60a5fa (product), teal #2dd4bf (email), orange #f59e0b (event), grey #6b6b6b (outbound)Typography: system-ui / sans-serif. Section labels: 11px uppercase 0.08em tracking. Body: 13–14px.
Layout: Single-column scroll. Cards with border-radius: 12px. Max-width 820px centered.
Header — Company initials avatar, company name, tagline (industry · HQ · headcount),
status tags (ICP tier from customer_fit_score, Intent level from engagement.intent_level),
last updated timestamp. In the top-right of the header, render a "View account on reo →"
link (anchor with target="_blank") pointing to firmographic.reo_account_link. Style it as
a small pill-shaped link: #2d2d2d background, #60a5fa text, border-radius: 999px, 11px
uppercase label with 0.08em tracking, subtle border 1px solid #3a3a3a, and hover state that
lifts to #333. If reo_account_link is missing or empty, omit the link entirely (do not
render a broken/placeholder button).
Overview — Row 1: Industry, Tech headcount (firmographic.tech_headcount), Founded Year,
HQ. Row 2: Tech stack badges from firmographic.technology_names. Missing fields show
"— Not available" in muted grey.
Scores — Two-panel row:
customer_fit_score with supporting signals listed
(industry match, tech headcount, hiring shape, dev stage)firmographic.developer_activity_score (0–100) with a
filled bar, plus Intent level badge from engagement.intent_level and Dev Stage from
firmographic.developer_stageActivity Summary — Chronological feed from top_activities[]: colored dot, actor name,
activity type, page label, source, date. Below: "What this suggests" inference block.
Cold badge if empty.
Developer Champions — From developers[]: avatar initials, name, email, title, activity
summary (counts by type), notable pages, first/last seen. Below all cards: "What this tells
us" inference block. Cold badge if none.
More Prospects — From prospects[]: avatar initials, name, title, location, LinkedIn URL
(linked), outreach note.
Hiring Signals — hiring.total_open_roles as a big number, tech function badges with
counts from hiring.by_tech_function[], insight banner inferring what the hiring shape means.
News & Trigger Events — From web search. Icon, title, implication, date. Omit if empty.
Competitive Context — From web search. Competitor cards with status. Objection prep if applicable.
Buyer KPIs by Role — 2×2 grid: role label, KPI list.
Suggested Next Steps — 3–5 action cards with urgency badges.
Response field reference — reo_get_account_research uses these field paths:
meta.org_id — UUID for the accountfirmographic.company_name, .industry, .hq.city/.state/.countryfirmographic.tech_headcount — integer, use for "Tech Headcount"firmographic.technology_names — array of strings, render as badgesfirmographic.customer_fit_score — string like "STRONG", "MODERATE" — drives the ICP sectionfirmographic.developer_stage — string like "HIGH", "LOW"firmographic.developer_activity_score — integer 0–100 — drives the Activity Score panelfirmographic.reo_account_link — URL for the "View account on reo →" header link (omit if missing)engagement.intent_level — "High", "Medium", "Low" — surfaced as a tag in the headerengagement.high_intent_hits[] — array of {page, date, actor}engagement.activity_type_breakdown — object with type→counttop_activities[] — the activity feed for the Activity Summary section (actor, type, page, source, timestamp)developers[] — {name, email, title, linkedin, github, location, first_seen, last_seen, signal_strength, activity_score, activity_summary, notable_pages, reo_link}prospects[] — {name, title, seniority, tech_function[], location, linkedin, email}hiring.total_open_roles, .by_tech_function[].function/.count, .date_rangeActivity signal hierarchy (for prioritising and narrating): FORM_CAPTURE > DEMO_LOGIN > COPY_TEXT/COPY_COMMAND > PAGE_VISIT > Identity
COPY_TEXT and COPY_COMMAND both indicate someone copied something — these are implementation signals stronger than PAGE_VISIT. Treat them as active evaluation indicators.