factcircuit
FactCircuit는 불변 출처와 단계별 검증, 제한된 검색, 재현 가능한 평가를 제공하는 감사 가능한 주장·증거 검증 루프입니다
프랙셔널 임원 시장이 100점 만점에 39.3점(냉각)이며, 30일간 10.7포인트 하락했습니다
The fractional executive market is at 39.3 of 100 (Cooling), down 10.7 points over 30 days. A private composite using 21 tracked inputs across demand, supply, culture, and context. SEC Form D filings
데이터TypeScript · JavaScriptGitHub @krishanraja온라인운영자 미확인
써 보셨어요? 한 번만 눌러 주세요. 로그인은 필요 없어요.
아직 감지된 업데이트가 없습니다. GitHub 릴리스가 나오면 여기에 보입니다.
개인 계정의 공개 저장소 소유자는 GitHub로 로그인해 운영자를 확인하고 소개를 직접 관리할 수 있습니다. 관리자 승인은 필요하지 않습니다.
로그인한 뒤 확인 파일을 만들어 이 저장소의 기본 브랜치에 30분 안에 커밋해 주세요. 다른 브랜치나 포크의 파일은 인정하지 않습니다. 파일을 올린 뒤 아래에서 확인하면 관리자 승인 없이 관리할 수 있습니다. 자격을 유지하려면 파일을 남겨 두세요.
로그인이 필요하면 로그인 화면으로 이동합니다. 로그인 후 이곳으로 돌아와 원하는 확인 버튼을 다시 눌러 주세요.
프로젝트 폴더에서 /nomorevibe verify를 실행하면 소유권을 확인하고 소개를 직접 관리할 수 있습니다.
1분 만에 확인하기The fractional executive market is at 39.3 of 100 (Cooling), down 10.7 points over 30 days. A private composite using 21 tracked inputs across demand, supply, culture, and context. SEC Form D filings provide financing context, not a validated forecast.
저장소 README에서 가져옴
README 전체를 불러오는 중입니다. 아래는 저장된 발췌입니다.
Pulse is the independent market instrument for fractional leadership. It publishes the Fractional Working Index (FWI), a free 0 to 100 composite showing whether demand is expanding, which C-suite roles are moving, and how strong the evidence is.
Pulse covers fractional CFO, CMO, CTO, COO, CRO, and interim CEO roles. Current role-demand collectors are US-scoped; English-language cultural inputs can include UK material but do not create a UK market benchmark. Twenty tracked inputs contribute to three pillars:README 전문 보기 ↗Pulse is the independent market instrument for fractional leadership. It publishes the Fractional Working Index (FWI), a free 0 to 100 composite showing whether demand is expanding, which C-suite roles are moving, and how strong the evidence is.
Live product · Current FWI JSON · Corporate strategy · Technical specification
Pulse covers fractional CFO, CMO, CTO, COO, CRO, and interim CEO roles. Current role-demand collectors are US-scoped; English-language cultural inputs can include UK material but do not create a UK market benchmark. Twenty tracked inputs contribute to three pillars:
FWI = (Demand × 0.50) + (Supply × 0.20) + (Culture × 0.30)
Pulse recalculates after each successful scheduled ingest. DataForSEO job tasks are prepared at 05:00 UTC, Vercel calls the main ingest at 06:00 UTC daily, and Monday has redundant ingest coverage. Weekly describes the briefing cycle and the seven-day role windows, not the score calculation frequency. Pulse is a private composite published by Fractionl, not an official or government index. It is not real-time, peer-reviewed, academically validated, backtested, or a promise of future market performance.
Fractional executives are an important audience, but they are not the primary payer. The public index, role pages, brief, REST API, and MCP tools remain free.
The first commercial hypothesis is a specialist fractional-executive marketplace, staffing or placement firm, or collective in the United States or United Kingdom that:
The economic buyer is usually the founder, CEO, managing director, or COO. They are not buying a larger quota for the public score. The proposed paid value is a private comparison of their book against a verified, multi-partner cohort.
Pulse is not an adtech product and is not for publishers as a paid product. Advertising buyers, media outlets, journalists, individual executives, single-hire companies, generic staffing firms, and AI-agent builders may use the public instrument. They are not the primary economic buyer.
The fractional leadership market still runs on anecdotes.
Specialist firms see their own pipeline but struggle to separate company-specific performance from market movement. Pricing, supply allocation, expansion, and client advice are often based on a small internal sample, public job searches, peer conversations, and generic staffing reports.
Public promise:
Read the fractional leadership market like an instrument.
Partner promise:
Benchmark your book against the market before your pipeline tells you too late.
Go Fractional already publishes free role, demand, and compensation benchmarks. Pulse must not claim to be the first or only fractional index. The paid thesis becomes differentiated only when Pulse combines the transparent public instrument with privacy-safe first-party engagement data from several independent partners.
Paid prices are strategic hypotheses until customers pay. Conditional plans are not generally available.
| Offer | Price | Status |
|---|---|---|
| Public instrument | £0 | Live, including score, history, role pages, brief, REST API, MCP tools, methodology, and source health |
| Founding Benchmark Partner | £1,500 for 90 days | Application-only validation offer, maximum ten organisations, subject to data and privacy review |
| Benchmark Membership | £6,000 per year | Conditional on ten active partners, 500 verified trailing-12-month engagements, and safe cohorts |
| Enterprise and portfolio | From £15,000 per year | Conditional on 1,500 verified engagements and sufficient cohort coverage |
Every displayed partner cohort must meet the approved minimum record count and contain data from at least three independent organisations. Exact privacy and suppression rules must be approved before partner-data ingestion launches.
See docs/CORPORATE_STRATEGY.md for the evidence matrix, assumptions, ICP, go-to-market plan, and validation gates.
Historical coverage is not uniform. Supply and culture do not have live coverage across the entire displayed period. Role pages measure demand, not role-specific supply.
Public reads require no authentication:
curl https://dtlcprcpvdomrehbejhw.supabase.co/functions/v1/fwi-api/current
curl "https://dtlcprcpvdomrehbejhw.supabase.co/functions/v1/fwi-api/history?months=3"
curl https://dtlcprcpvdomrehbejhw.supabase.co/functions/v1/export-brief
Base URL: https://dtlcprcpvdomrehbejhw.supabase.co/functions/v1
| Endpoint | Purpose |
|---|---|
GET /fwi-api/current | Latest composite, components, prior-period change, movers, source details, and completeness |
GET /fwi-api/history?months=N | Historical score observations, with N clamped to 1 through 12 |
GET /fwi-roles | Demand readings for the six covered roles |
GET /export-brief | Weekly brief as Markdown or JSON |
POST /fwi-api/trigger | Service-role-only ingestion trigger, not public |
Optional API keys support operational rate controls. They are not a paid generic-data tier.
Hosted MCP server:
https://dtlcprcpvdomrehbejhw.supabase.co/functions/v1/mcp
Tools:
get_fractional_working_indexget_fwi_weekly_briefget_content_radarget_content_briefDiscovery surfaces:
/product-truth.json/llms.txt/.well-known/ai-plugin.json/sitemap.xml/feed.xmlThe weighting is Demand 50%, Supply 20%, and Culture 30%. If a pillar has no live data for a reading, its weight is redistributed proportionally and the response exposes the effective weights and completeness.
The tracked inputs include job postings, search interest, company-financing context, professional profiles, GoFractional's published operator-network size, public discourse, news, podcasts, self-employment, and macro context. Context signals enrich interpretation but are excluded from the composite where documented.
Integrity controls include:
The SEC Form D input is treated as a financing-context signal that may precede some fractional hiring. Its predictive relationship has not been validated, so Pulse must not attach a forecast-accuracy claim to it.
| Layer | Stack |
|---|---|
| Frontend | React 18, TypeScript, Vite, Tailwind CSS, shadcn/ui, Framer Motion, Recharts, React Query |
| Backend | Supabase Postgres and Edge Functions |
| Auth | Supabase Auth |
| AI summaries | OpenAI model invoked from server-side functions, with caching and rate controls |
| Scheduling | Supabase pg_cron prepares DataForSEO Jobs at 05:00 UTC; Vercel schedules the main ingest at 06:00 UTC daily and a Monday backstop; Supabase retains a Monday ingest backstop |
| Deployment | Vercel at pulse.fractionl.ai |
Core Edge Functions:
| Function | Responsibility |
|---|---|
prepare-dataforseo-jobs | Submit six idempotent async DataForSEO Google Jobs tasks at 05:00 UTC, record task IDs, and leave retrieval to the 06:00 ingest |
ingest-signals | Collect and normalise tracked inputs |
calculate-fwi | Produce the composite and movers |
fwi-api | Serve current, historical, and protected trigger routes |
fwi-roles | Serve role-demand readings |
export-brief | Produce the weekly market brief |
generate-pulse-insights | Produce cached summaries grounded in the index pipeline |
fwi-verdict | Answer a scoped index question with origin, rate, and prompt controls |
mcp | Expose the hosted MCP server |
send-pipeline-alert | Send operational alerts |
Requirements: current Node.js LTS and npm.
git clone https://github.com/krishanraja/fractionl-pulse
cd fractionl-pulse
npm install
Copy-Item .env.example .env
npm run dev
The browser bundle requires:
VITE_SUPABASE_URL
VITE_SUPABASE_ANON_KEY
Server routes and scheduled jobs use:
SUPABASE_URL
SUPABASE_SERVICE_ROLE_KEY
CRON_SECRET
See .env.example for the complete integration list. Never commit private keys or service-role credentials.
npx tsc --noEmit
npm run build
npm run lint
npm run docs:audit
npm run audit
npm run audit:json
npm run smoke:release -- --skip-site
The audit is read-only. It checks schedule adherence, source health, completeness, and provenance against the committed baseline. A degraded audit is an incident to investigate, not a result to hide. smoke:release exercises the public APIs, MCP compatibility, security boundaries, Ask the Index stream, discovery files, and live site routes; set PULSE_URL to verify a preview before production.
| Document | Purpose |
|---|---|
| Agent entry file | What a coding agent reads first: routes to NOW.md, then this repository's own rules. Carries the marker-delimited canon block rendered from krishanraja/ai-harness |
| Where it is right now | Current-state router: lifecycle, the main commit it was reconciled against, recent changes, what to read next, and what not to trust |
| History log | Append-only record of reconciliations, decisions, and superseded documents moved into docs/history/ |
| Corporate strategy | Canonical ICP, buyer pains, positioning, pricing, evidence, and validation gates |
| North star | User outcome, buyer outcome, moat metric, and near-term gate |
| Monetization strategy | Packaging, price tests, and commercial rules |
| Sales playbook | Qualification, discovery, pilot, objections, and outreach |
| Agent briefing | Current claims, offers, and prohibited language for AI agents |
| Autonomous GTM playbook |
| Research, scoring, messaging, qualification, and proposal rules for autonomous marketing and sales agents |
| Documentation governance | Truth hierarchy, status labels, ownership, and reconciliation rules |
| Design system | Information architecture, branding, device-specific layouts, interaction rules, and release acceptance widths |
| Technical specification | Architecture, schema, functions, and calculations |
| Agent integration | REST response and integration reference |
| MCP tool guide | MCP tools and implementation examples |
| Data-sources roadmap | Input inventory and future data work |
| Weekly pipeline audit | Operational audit runbook |
| Answer hub | Question-led answer pages at /answers: cadence, sourcing, quality bar, how to add a post |