Algorithmic Trading Firm Licensing Thresholds logo

Algorithmic Trading Firm Licensing Thresholds

Community
HimanshuJ16
algorithmic-trading-firm-licensing-thresholds

Use when auditing whether a proprietary trading firm has crossed a quantitative registration trigger: the 17 CFR 240.15b9-1 conditions for FINRA membership exemption, the MiFID II Article 19 message-rate test, and the SEBI limbs.

Overview

PublisherHimanshuJ16
RepositoryAlgo-Trading-Skills
Skill namealgorithmic-trading-firm-licensing-thresholds
Stars
50
Forks
7
Bundled files
5
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by HimanshuJ16 on GitHub. Read the source before you install it.

Installation

Install the Algorithmic Trading Firm Licensing Thresholds AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/HimanshuJ16/Algo-Trading-Skills/tree/main/skills/algorithmic-trading-firm-licensing-thresholds
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/HimanshuJ16/Algo-Trading-Skills.git /tmp/Algo-Trading-Skills
mkdir -p .claude/skills
cp -r /tmp/Algo-Trading-Skills/skills/algorithmic-trading-firm-licensing-thresholds .claude/skills/algorithmic-trading-firm-licensing-thresholds
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Algorithmic Trading Firm Licensing Thresholds in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Algorithmic Trading Firm Licensing Thresholds on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Algorithmic Trading Firm Licensing Thresholds is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

When to Use

Use this skill when auditing a proprietary trading firm, hedge fund or trading member to check whether its activity has crossed a quantitative registration trigger in the US, the EU or India — the numeric limbs of three specific regimes, screened against the numbers the regulator or the exchange actually publishes.

Start here by correcting the misconception this skill exists to prevent. There is no single "algo trading firm licence" and no global message-rate threshold that triggers one. The three regimes look superficially alike and are not:

  • US. Rule 15b9-1 is an exemption from the Exchange Act section 15(b)(8) requirement to join FINRA, not from broker-dealer registration under section 15(a), and nothing in it turns on order or message rates. It turns on three conditions: exchange membership, carrying no customer accounts, and executing solely on an exchange of which the firm is a member.
  • EU. The MiFID II HFT designation is a message-rate test — but on an average, at 2 and 4 messages per second, not on a peak and not at any round number in the dozens. Meeting it removes the own-account dealing exemption and forces investment firm authorisation.
  • India. The SEBI Threshold Orders Per Second registers an algorithm with an exchange, for a retail investor's API-routed flow. It is not entity licensing and it does not govern a trading member's own flow.

The output is an auditable LicensingComplianceReport with three outcomes, not two: a crossed threshold (requires_registration), an undetermined question (manual_review_required), and neither. Any report that is not is_clear goes to the Chief Compliance Officer and qualified regulatory counsel before the desk continues operating.

When NOT to Use

  • As a licensing determination. The engine screens numeric limbs. Whether a firm is a "dealer" or is "engaged in the business" at all, whether an EU activity is an investment service, and every qualitative fact besides, sit outside it. A clear report is not a legal opinion.
  • As evidence a registration exists. The engine reads flags the caller supplies. It cannot confirm a firm is a FINRA member, holds a MiFID authorisation, or has an algorithm registered with an exchange.
  • For an Indian trading member's own algorithms. TOPS governs retail API-routed flow. The exchange algo-approval regime that governs a member's own algorithms is not modelled; the IN branch says so rather than returning a clean report — see india-sebi-algo-trading-tagging-requirements.
  • For a US message-rate trigger. There isn't one in Rule 15b9-1. If you need a US order-rate control, that is a pre-trade risk obligation — see sec-rule-15c3-5-risk-controls-us.
  • Outside US/EU/IN. Unknown jurisdictions fail closed to manual review. Do not carry the 2/4 messages-per-second figures or the 10 OPS TOPS into another jurisdiction; see mas-singapore-algo-trading-guidelines for a worked example of how differently the same-sounding obligations are built.

Prerequisites

  • Python 3.10+ (standard library only).
  • A FirmTradingActivity snapshot over a documented evaluation window:
    • jurisdiction, is_exchange_member, has_customers.
    • off_exchange_volume_usd and exempt_off_exchange_volume_usd — the latter being the portion falling solely within the Rule 15b9-1(c) exceptions, and evidenced, not assumed.
    • peak_orders_per_second — the highest order count in any single calendar clock second, per exchange, which is the basis NSE specifies for TOPS. Not a rolling window.
    • For the EU: avg_messages_per_second_per_instrument and avg_messages_per_second_all_instruments, computed per Article 19(1) and restricted to liquid instruments per Article 19(2). Leave them None if unmeasured — None and 0.0 mean different things here.
    • is_retail_api_algo_flow for the IN branch.

The module computes none of these from raw order data; aggregation discipline (windowing, exchange-versus-ATS classification, Article 19(2) message exclusions) is a prerequisite, not a feature.

Workflow

  1. Aggregate, then construct. Build FirmTradingActivity. The constructor rejects unsupported jurisdictions, non-finite and negative metrics, bool masquerading as a numeric quantity, and an exempt off-exchange figure larger than the total — a claim that would otherwise net to a negative non-exempt volume and hide a condition (c) breach.
  2. Configure only to tighten. LicensingThresholdEvaluator defaults to the published figures. Override sebi_tops_orders_per_second, mifid_ii_msgs_per_sec_single_instrument, mifid_ii_msgs_per_sec_all_instruments or sec_off_exchange_floor_usd only to screen more strictly. Loosening one past the published figure puts the firm outside the rule it is screening for. 0 is a valid, stricter override and is honoured as such.
  3. Evaluate. evaluator.evaluate(activity) runs every check; nothing short-circuits. The customer-account check runs first and dominates rule_id, but never suppresses the jurisdiction's own violations.
  4. Triage on three outcomes, not two.
    • requires_registration — a modelled threshold was crossed. Throttle, disable the offending routing, or stop, and escalate.
    • manual_review_required — the evaluator could not conclude. Escalate to counsel. Do not record this as compliant, and do not act on it as though a breach were confirmed either. The distinction matters: an EU firm whose Article 19 averages were never computed is an open question, not a proven HFT.
    • is_clear — neither fired. Still not a legal opinion.
  5. Audit. Persist the report (evaluated_at, schema_version, rule_id, violations, manual_review_items) alongside the input snapshot. Re-run on a documented cadence; ESMA expects a firm to self-assess its Article 19 position at least monthly.

Common Pitfalls

  • Reading Rule 15b9-1 as a registration exemption. It exempts a broker-dealer from joining FINRA under section 15(b)(8). A firm that concludes it need not register as a broker-dealer because it fits 15b9-1 has read the wrong statute.
  • Screening EU exposure on an order rate. Article 19 is measured on an average, and it counts messages — modifications and cancellations included. One order per second, cancel-replaced five times, is already above the 2 messages/second limb. So peak_orders_per_second is not an EU input in either direction: a low order rate does not earn a clean report, and a high one does not prove a breach. Without both averages the engine returns "cannot determine" — where the earlier 50-peak-OPS benchmark returned "compliant" for firms comfortably inside the HFT definition.
  • Treating a missing measurement as a zero. None for an Article 19 average means unmeasured. Passing 0.0 asserts you measured zero traffic, and will produce a clean report you cannot defend.
  • Flagging all off-exchange volume. Rule 15b9-1(c) still permits exchange-routed Rule 611 / Options OPP flow and the stock leg of a stock-option order. Netting those out is what exempt_off_exchange_volume_usd is for — but (c)(2) requires written policies and procedures preserved for three years, so the engine raises a review item whenever the exception is claimed rather than accepting it.
  • Assuming a de minimis allowance survives. It does not. The 2023 amendments removed it; the default screening floor is therefore 0.00 USD. A higher floor is a firm's own triage threshold and reflects no regulatory carve-out.
  • Applying TOPS to a proprietary desk. It governs a retail investor's API-routed algorithm, and requires registering the algorithm with each exchange through the broker — not licensing the firm.
  • Missing the TOPS boundary. NSE sets it at "not exceeding 10 orders per second per exchange", so registration bites above 10, not at 10.
  • Logging user-provided fields verbatim. Build downstream logging on report attributes with %s placeholders, never on raw call-site strings, to avoid log injection through free-form fields.

Verification

Run python -m unittest discover -s skills/algorithmic-trading-firm-licensing-thresholds/scripts (58 tests). The suite asserts each jurisdiction's threshold at and around its boundary, that the Article 19 limbs fire independently at 2.0 and 4.0 messages/second, that neither a low nor a high order rate can decide the EU question, that a 0 threshold override is honoured rather than silently replaced by the class default, that bool is rejected wherever a numeric quantity is expected, that exempt off-exchange volume is netted out — without a sub-cent floating-point residue reading as a breach — and its evidencing obligation surfaced, that violations preserve evaluation order rather than sorted order, and that an unrecognised jurisdiction fails closed to manual review. Verify the outcome against assets/checklist.md.

Related Skills

  • finra-algo-trading-registration-requirements
  • sec-rule-15c3-5-risk-controls-us
  • mifid-ii-algo-trading-compliance-eu
  • india-sebi-algo-trading-tagging-requirements
  • mas-singapore-algo-trading-guidelines

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Algorithmic Trading Firm Licensing Thresholds AI skill do?

Use when auditing whether a proprietary trading firm has crossed a quantitative registration trigger: the 17 CFR 240.15b9-1 conditions for FINRA membership exemption, the MiFID II Article 19 message-rate test, and the SEBI limbs.

Why use Algorithmic Trading Firm Licensing Thresholds on TypingMind?

Because you install it once and use it with any model. Algorithmic Trading Firm Licensing Thresholds is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Algorithmic Trading Firm Licensing Thresholds in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/HimanshuJ16/Algo-Trading-Skills/tree/main/skills/algorithmic-trading-firm-licensing-thresholds. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Algorithmic Trading Firm Licensing Thresholds?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Algorithmic Trading Firm Licensing Thresholds?

As many as you like. As long as a model supports skills, you can use Algorithmic Trading Firm Licensing Thresholds with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Algorithmic Trading Firm Licensing Thresholds AI skill free?

Yes. It is published on GitHub by HimanshuJ16 under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇