FAQ

Everything people actually ask.

Compliance, methodology, pricing, and what it takes to run a system like this yourself.

Research & education only. Not investment advice. · WATCHLIST_ONLY

FAQ

Questions, answered plainly.

What is TradingMapClaw?
TradingMapClaw (TMC) sells original digital education products and AI research-workflow templates, built around a dual-engine, multi-model verification workflow. It runs 118 scheduled jobs on a single Mac mini, cross-verifies its own outputs with a council of models, and documents everything so builders can rebuild it themselves. It is education and research only — WATCHLIST_ONLY, with no broker API and no order execution. Nothing it produces is investment advice.
How is it different from asking ChatGPT?
A single model gives one answer with no second opinion — it can be confidently wrong and nothing tells you. TMC uses two independent engines: Engine 1 (Hermes Agent) derives the fundamentals, and Engine 2 (Codex) independently re-derives the key numbers. If they agree within 5%, confidence rises; if they disagree, the divergence is printed, not smoothed over. Contested calls go to a multi-model Council War Room vote.
How much does it cost to run?
The system runs under a $55/month headline budget cap. Actual spend is about $7/month — roughly 13.5% of the cap — thanks to a 4-tier model fallback chain that ends in a free local model (Qwen3 14B).
Is TradingMapClaw investment advice?
No. TMC produces research and analysis only. It does not route orders, execute trades, or provide personalized investment advice. All output is for research and educational purposes and is not investment advice. It operates in WATCHLIST_ONLY mode by design.
What models does it use?
The council includes DeepSeek V4 Pro, GLM-5.2, and GPT-5.6, plus a local Qwen3 14B. The fallback chain runs GLM-5.2 → GPT-5.6 → DeepSeek V4 Pro → Qwen3 14B (local, free).
What are the skill packs?
Seven standalone engineering patterns distilled from 50+ internal SKILL modules: Budget Watchdog, Maker-Checker Cross-Verification, Cron Recovery / Self-Heal, Quality Gate, Model Fallback Chain, Prompt Governance, and Generic Data Fallback Chain. All are environment-agnostic and OpenAI-compatible. The open-source example code is MIT-licensed; paid product materials are for personal and internal use only and may not be resold or redistributed.

Practical

Getting started.

How is this different from Seeking Alpha or Koyfin?

Those are excellent content and data subscriptions — you rent access to someone else's analysis or terminal. TradingMapClaw teaches you to own the machine: your own pipeline, your own reports, your own cost structure (~$55/mo), fully auditable and yours forever. See the comparison below.

What do I need to run the system?

A Mac or Linux machine (a base Mac mini is plenty), Python 3.10+, free-tier API keys for market data, and roughly $55/month for AI model usage at full scale — far less while you're learning.

Can you set it up for me?

That's what Operator Consulting is for — a recorded 60-minute session on your architecture, plus a written action plan. We advise on the workflow; you keep the keys.

What's the refund policy?

Digital downloads are delivered instantly, so sales are generally final — but corrupted, incomplete, or duplicate orders are refunded or replaced (email us within 14 days). Consulting: free reschedule or full refund up to 24h before the session. Full details in the refund policy.

Can I see the quality before paying?

Yes — the Radar page shows redacted sample output, and the GitHub repo documents the architecture publicly.

How It Compares

Against every alternative — and honest about the trade-offs.

Commercial platforms, open-source frameworks, and the DIY ChatGPT-and-spreadsheet approach. Here's where TradingMapClaw actually stands.

Capability TradingMapClaw Commercial platforms
Bloomberg · AlphaSense		 Open-source frameworks
OpenBB · TradingAgents		 Individual builders
Solo scripts		 DIY
ChatGPT + Excel
Annual cost~$660$18k–$32kFree (self-run)Varies~$240
Transparent & auditableYesBlack boxYesDependsPartial
Education & research (not trade signals)Education-firstYesSignal / sim-executionMixedYes
Customizable to your watchlistFullyLimitedYes (code)YesManual
Automated scheduling118 scheduled jobsYesDIY setupDependsNo
Dual-engine quality gateMaker/CheckerNoNoNoNo
Compliance guardrailWATCHLIST_ONLYVendor termsNoneNoneNone
Bilingual EN + ZH deliveryBuilt-inNoNoNoNo
Solo-operableBy designTeam-orientedDev effortYesYes
Coverage scale82 tickersUniverseConfigSmallHandful
Reproducible from docsDocumentedNoYesRarelyPartial

Commercial pricing reflects published list rates (Bloomberg Terminal ~$31,980/yr; AlphaSense enterprise median ~$18,375/yr). Open-source stars: OpenBB, TradingAgents. TradingMapClaw's edge is not raw scale — it is transparency, an architecture-level compliance boundary, a dual-engine quality gate, and bilingual delivery, all under a hard $55/month ceiling. It is deliberately not a trading bot, and nothing it produces is investment advice.

$660
TradingMapClaw
per year
$2,400
Seeking Alpha Pro
per year
$18,375
AlphaSense
per year (median)
$31,980
Bloomberg Terminal
per year

WATCHLIST_ONLY: TradingMapClaw produces education and research materials only. It does not route orders or execute trades, and nothing it produces is investment advice — that boundary is intentional.