Hypergrok Bootstrap logo

Hypergrok Bootstrap

Organization
galleonlabs
hypergrok-bootstrap

Build and verify a HyperGrok trading desk from the pinned public release. Use for first-run setup, repair, or a readiness check. Starts with a live zero-key Opening Bell, installs the seven role profiles and seventeen shared skills, prepares the Trading Floor, and returns an evidence receipt. Read-only by default; never requests a wallet or places an order.

Overview

Publishergalleonlabs
Repositoryhypergrok-trading-desk
Skill namehypergrok-bootstrap
Stars
145
Forks
31
Bundled files
1
LicenseMIT
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.

  • 1 bundled files

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

  • Open source

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

Installation

Install the Hypergrok Bootstrap 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/galleonlabs/hypergrok-trading-desk/tree/main/skills/hypergrok-bootstrap
  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/galleonlabs/hypergrok-trading-desk.git /tmp/hypergrok-trading-desk
mkdir -p .claude/skills
cp -r /tmp/hypergrok-trading-desk/skills/hypergrok-bootstrap .claude/skills/hypergrok-bootstrap
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Hypergrok Bootstrap 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 Hypergrok Bootstrap 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 Hypergrok Bootstrap 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.

HyperGrok bootstrap

Turn a fresh shared Desk Lead into a working HyperGrok desk. Finish with evidence, not a claim that setup probably worked.

Safety boundary

Bootstrap is research-only.

  • Do not request, read or store an API wallet, private key, seed phrase or exchange secret.
  • Do not call Hyperliquid /exchange, create an order, alter leverage, move funds or change approval settings.
  • Public POST /info market reads are allowed. State mainnet, request type and UTC time.
  • Local writes are limited to /workspace/hypergrok and /workspace/trading-desk.
  • Creating the seven named Bots and the private Trading Floor group is in scope. Sharing anything publicly is not.
  • If a capability is unavailable, return the exact manual card or step and continue independent supported work. Do not pretend it happened or mark dependent checks complete.

1. Install the reviewed release

If /workspace/hypergrok is already a Git checkout, preserve it. Verify that plugin.json names version 1.6.1 and that git rev-parse HEAD equals git rev-parse 'refs/tags/v1.6.1^{commit}'. If the tag is unavailable, the version differs or the commits differ, report the existing version and commit and stop dependent setup. A structural pass from an older release is insufficient; do not overwrite, reset or switch the existing checkout. Otherwise clone the pinned release:

bash
mkdir -p /workspace && cd /workspace
git clone --depth 1 --branch v1.6.1 https://github.com/galleonlabs/hypergrok-trading-desk.git hypergrok
cd /workspace/hypergrok
git rev-parse HEAD
bash scripts/check.sh

Run bash scripts/check.sh for an existing verified checkout too. Stop if the structural check fails. Record the printed commit for desk.md.

Before using any agent profile or System prompt, run git diff --exit-code refs/tags/v1.6.1 -- agents/ in the checkout. A matching HEAD can still have changed working files. If the command reports differences or fails, name the affected paths, stop dependent setup and preserve the user's files.

2. Ring the Opening Bell

Show useful output before asking the user configuration questions:

bash
cd /workspace/hypergrok
python3 scripts/opening_bell.py --coin ETH

Return the output verbatim. It is a public market snapshot, not a signal. If it is unavailable, say so and continue offline; do not substitute stale or invented figures.

3. Prepare the desk

bash
mkdir -p /workspace/trading-desk/{proposals,briefs,research,strategies,data,journal/incidents,watch}
cd /workspace/hypergrok
python3 scripts/desk_doctor.py --desk-root /workspace/trading-desk

A warning for the not-yet-written desk record is expected at this stage. A failed repository or public API check is not.

4. Build the team

Read docs/ARCHITECTURE.md, skills/desk-operating-model/SKILL.md, every file in agents/, and skills/README.md.

Select one Bot for each agent file's Bot profile and System prompt:

  1. Desk Lead
  2. Market Analyst
  3. Research Analyst
  4. Strategist
  5. Risk Manager
  6. Execution Trader
  7. Trade Reviewer

Read back each selected Bot's native Name, Job and Description and compare them exactly with its profile. Verify its standing instructions before reuse or after creation. Supported evidence is either the complete stored System prompt matching this release, or readback of the exact full System prompt delivered as a standing-instructions message, the Bot's acknowledgement that it follows the pinned agent file, and that local file's hash matching the expected bytes from git show 'refs/tags/v1.6.1:agents/<name>.md'. Hashing an arbitrary local file is insufficient. Label the latter delivery-and-pointer: it proves delivery and the checked target, not byte-identical persistent memory. A matching Description, summary or path alone is insufficient.

If an existing seat differs, is missing or cannot be verified, create a separate Bot from the reviewed profile and deliver its full System prompt as directed in SETUP.md section 4. Preserve existing Bots; reuse does not authorize replacing an unrelated desk's profiles, prompts or memories. If a needed creation step or both supported verification methods are unavailable, return that role's exact manual card and keep its readiness pending. Record the native Bot ID for every verified seat.

Use assets/mascot.jpg as the avatar when the product supports it. Do not add unrelated memories, conversation history or private files. Continue independent local preparation and checks; keep Bot and group readiness pending until existence and instruction evidence are verified.

Inspect the shared skills before installing anything. A Desk Lead added from the public HyperGrok template carries the reviewed bootstrap skill. Install the remaining release skills from this checkout; do not create duplicate skills. Verify actual native shared instructions before recording them as template; that status means an already-present verified skill, not that all seventeen were exported by the public share.

Use supported native readback to inspect each skill's exact name, exact description and complete Markdown body. Grok stores name and description separately and rebuilds the YAML wrapper without source license or metadata. Compare the body after removing frontmatter, leading empty lines and the final newline only; preserve all internal text, spaces and line breaks. Do not strip trailing spaces from internal lines or summarize content. Full-file hashes in template/grok-bot.json continue to verify all seventeen pinned checkout files. Checkout hashes and staged save arguments do not prove what the native product stored.

For each of the seventeen skills/*/SKILL.md files:

  1. If an existing shared skill passes complete native readback against the release, enable it for the selected Desk Lead and record template.
  2. If it is missing, save the exact source name, description and normalized body. Read the saved skill back and record installed only when they match.
  3. If the product cannot save that body, save the source name and description with a pointer instructing it to read /workspace/hypergrok/skills/<name>/SKILL.md. Verify the complete stored pointer text and the target's full-file hash, then record pointer.
  4. If native content differs or cannot be read back completely, record mismatch with the reason and fail native readiness. Do not overwrite shared content used by unrelated desks without user authority. Never claim a skill is current from its name, a local file or a staged argument alone.

The final receipt must contain exactly seventeen unique skill names and one status for each: template, installed, pointer or mismatch.

Create a private group chat named Trading Floor with the six selected Bot IDs for Desk Lead, Market Analyst, Research Analyst, Strategist, Risk Manager and Execution Trader. Reuse an existing group only when its ID is already selected for this desk and native readback confirms exactly those member IDs; its name alone is insufficient. Preserve prior groups. Record the Floor ID and member IDs and direct the welcome message and role checks to that group. The Trade Reviewer stays in DM. If group creation or membership readback is unavailable, return the exact manual step and keep group readiness pending.

5. Record the research desk

Unless the user explicitly selects testnet or mainnet, create a research desk:

Record only verified Bot, group and member IDs as observed; mark uncreated, manual or unverified roles and any unverified Floor as pending rather than copying the desired seven-Bot/six-member inventory as an observed result.

markdown
# Desk record

- created: <UTC time>
- instructions commit: <commit from step 1>
- engagement level: research
- network: mainnet-for-reads
- account: none
- bots: Desk Lead, Market Analyst, Research Analyst, Strategist, Risk Manager, Execution Trader, Trade Reviewer
- bot ids: <role to verified native Bot ID>
- group chats: Trading Floor (6)
- trading floor id and member ids: <verified native group and six Bot IDs>
- risk limits: not yet written
- standing approvals: none
- exchange approval gate: unverified; required before provisioning any API wallet key
- unprotected position deadline: not applicable
- status: research-only; no API wallet provisioned

Write it to /workspace/trading-desk/desk.md. Do not provision trading access during bootstrap.

6. Verify and return the receipt

Run:

bash
cd /workspace/hypergrok
python3 scripts/desk_doctor.py --desk-root /workspace/trading-desk
python3 scripts/opening_bell.py --coin BTC

Then perform the five role checks in SETUP.md section 9. Return:

  • release tag and installed commit
  • Opening Bell source, network and UTC time
  • each Bot: ID, created/reused/manual card, profile comparison and standing-instruction proof method; state any delivery-and-pointer limitation
  • each skill: status, native readback method or unavailable reason, normalized stored body length/hash and comparison result; identify pointer evidence separately
  • Trading Floor ID and verified member IDs
  • desk doctor results, including warnings
  • five role-check results
  • exact local paths for desk.md and the repository
  • this sentence: Setup stayed read-only: no key requested and no order created or sent.

Do not say the desk is ready if a required check failed. Name the failed check and the next safe action.

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 Hypergrok Bootstrap AI skill do?

Build and verify a HyperGrok trading desk from the pinned public release. Use for first-run setup, repair, or a readiness check. Starts with a live zero-key Opening Bell, installs the seven role profiles and seventeen shared skills, prepares the Trading Floor, and returns an evidence receipt. Read-only by default; never requests a wallet or places an order.

Why use Hypergrok Bootstrap on TypingMind?

Because you install it once and use it with any model. Hypergrok Bootstrap 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 Hypergrok Bootstrap in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/galleonlabs/hypergrok-trading-desk/tree/main/skills/hypergrok-bootstrap. 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 Hypergrok Bootstrap?

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 Hypergrok Bootstrap?

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

Is the Hypergrok Bootstrap AI skill free?

Yes. It is published on GitHub by galleonlabs under the MIT 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 👇