← Back to Skills Marketplace
cryptorabble

Empire Builder

by CryptoRabble · GitHub ↗ · v1.3.0 · MIT-0
cross-platform ⚠ suspicious
143
Downloads
0
Stars
0
Active Installs
5
Versions
Install in OpenClaw
/install empire-builder
Description
Empire Builder AI skill — SmartVault treasuries, leaderboards, boosters, prepare→executeBatch→store distributions, burns, airdrops, Clanker deploys. Self-con...
README (SKILL.md)

Empire Builder Skill

Empires are community treasuries (ERC-4337 SmartVaults on Base 8453 and Arbitrum 42161) paired with an identity (Empire ID). Agents orchestrate HTTP APIs plus wallet-signed txs (executeBatch for payouts).

Base URL: https://empirebuilder.worldmainnet only. There is no testnet or sandbox API documented here.

Packages (pinned): npm install [email protected] [email protected] — versions match the YAML frontmatter; use a lockfile in automation; avoid @latest for reproducibility.

Skill copies: If you sync from homepage, diff against a trusted checkout before relying on it in scripts.

\x3Ca id="intentional-product-behavior">\x3C/a>

Intentional product behavior (not documentation bugs)

The integration surface is meant to hit production chains with vault owner() authority. “Hardening” belongs in product code, infra, and operator tooling—this skill only describes the API/contract shape.

  • All writes are live on Base 8453 / Arbitrum 42161 unless you stick to reads.
  • distribute-prepare + owner-broadcast executeBatch is the designed API path; SmartVault owner() is required on-chain (see references/contracts.md).
  • POST /api/store-distribution persists from mined transactionHashes (receipt-backed): recorded payouts follow on-chain execution, not prepare JSON alone. You still must pass correct empireAddress, baseToken (Empire ID), and leaderboard fields so rows attach to the right empire (http-api.md).
  • Registry/scanner warnings about “agent + wallet + mainnet” reflect that intentional design, not missing caveats in this file.

How agents should use this skill

Models usually load SKILL.md first only. Treat this file as routing + constraints. Before writing HTTP bodies, signing flows, or multi-step flows, open the reference files listed below—otherwise you will guess missing fields (signer vs signerAddress, leaderboardType values, booster payloads). For one pipeline story (especially payouts), follow references/workflows.md step order.


When NOT to use this skill

Stop and pick a different approach if any of these apply:

  • End-user wallet UX inside the Empire web app. The website signs sponsored ERC‑4337 UserOperations (paymaster-funded executeBatch via EntryPoint) for co-signers — not the prepare → owner-broadcast → store path this skill describes. Use the in-app flow (or its dedicated front-end client code), not these HTTP routes.
  • Co-signer / "authorized signer" automation. Co-signers cannot call vault execute/executeBatch directly (reverts Unauthorized) and cannot drive POST /api/distribute-prepare — that route requires owner(). If you are a co-signer, either (a) trigger a UserOp through the web app, or (b) escalate to the vault owner.
  • Generic ERC-4337 / paymaster / bundler tutorials. This skill does not teach you to build a sponsor, bundler, or EntryPoint integration. Don't graft /alchemy-sponsor-userop-style helpers or legacy distribute-v3 endpoints onto these flows.
  • Non-Empire token deploys. Plain Clanker SDK deploys without an Empire are out of scope — use the Clanker docs directly. This skill only covers Clanker → deploy-empire wiring.
  • Testnets / local dev. No sandbox endpoint exists. If you need a dry-run, read-only GETs are safe; do not invent /sepolia/..., chainId: 84532, or staging hosts.
  • Mass user-to-user transfers from random EOAs. Use a normal ERC-20 transfer from the holding wallet — vault roles, prepare/store, and distribution routes don't apply.

Empire ID vs SmartVault address (read this first)

Three input shapes resolve to one Empire, but they are not the SmartVault address:

Shape Looks like Where it appears
Token empire 0x + 40 hex (the ERC‑20 base token) most tokenAddress / baseToken params
Farcaster tokenless fid + digits (e.g. fid12345) tokenless empires anchored to a Farcaster fid
Custom tokenless slug alphanumeric slug tokenless empires with a custom identity
  • Empire ID is identity. Often spelled tokenAddress, baseToken, empire_id, or sits in the path segment /api/empires/[empire_id].
  • SmartVault address is the on-chain treasury contract. Spelled empire_address / empireAddress / treasuryAddress. Resolve it from GET /api/empires/[empire_id].
  • They are never interchangeable unless an endpoint explicitly takes both. Do not pass an fid… Empire ID where a contract address is required, and do not pass the SmartVault address where Empire ID is required (e.g. store-distribution wants both, with distinct field names).

See references/http-api.md → Empire ID for the full type table.


How this skill is structured

Everything agents need lives inside skill/ — no dependency on other folders.

File Role
SKILL.md (this file) Scope, rules, orientation
references/http-api.md Canonical HTTP routes, auth, Empire ID, key JSON shapes
references/workflows.md End-to-end sequences (deploy, distribute, burn, …)
references/contracts.md SmartVault execute / executeBatch, chains

Agent constraints (mandatory)

  1. Catalog fidelity. Only use routes and semantics described in references/http-api.md. Do not invent paths or composite undocumented URLs.

  2. Authentication. Send x-api-key where references/http-api.md says so. Writes often require EIP-191 signature, message, and signer or signerAddress — match field names per route.

  3. Treasury payouts. Integration flow only: POST /api/distribute-prepare → vault owner() wallet submits vault executeBatch txs (you pay gas) → POST /api/store-distribution. Do not substitute random sponsor/UserOp tutorials unless working purely inside the official web app (different pipeline).

  4. POST /api/distribute-prepare signer must equal owner() on the vault — not co-signers / “authorized” SmartVault signers. Those addresses can authorize UserOperations in the web app (sponsored executeBatch via EntryPoint), but they cannot use this API path and cannot execute or executeBatch on the vault with a normal EOA transaction — the contract reverts Unauthorized. See references/contracts.md. On other routes, recovered signer may be checked against owner / co_emperors per backend rules — do not assume the same rule as distribute-prepare.

  5. Empire ID vs SmartVault address. Per the dedicated section above, tokenAddress / baseToken / empire_id is identity (ERC‑20, fid…, or slug); empireAddress / treasuryAddress is the on-chain vault. Confusing them is the most common skill-failure mode — re-check field names against references/http-api.md before sending any write.

  6. Burns. Decoded from the burn tx receipt: ERC‑20 Transfer to 0x000…0000 or 0x000…dEaD for the empire base (or attached base if tokenless), then POST /api/store-burn. If tokens sit in the SmartVault, only owner() can fire a direct vault execute / executeBatch; co-signers burn from the treasury in the website via sponsored executeBatch UserOps (Base / Arbitrum). If tokens sit in any other wallet, that wallet **transfer**s to a burn address — no vault role needed.


At-a-glance: integration surfaces

Area Mechanism
Reads Mostly open GETs; some routes require API key (see references/http-api.md)
Leaderboard payouts prepare (owner-signed) → transactions[] → owner broadcasts executeBatch → store-distribution
Burns Eligible Transfer in mined tx → store-burn; treasury burns: owner self-paid execute, or web app UserOp executeBatch for co-signers
Boosters references/http-api.md/api/boosters/...
Token deploy get-token-config → Clanker SDK deploy → deploy-empire

Example prompts

  • Walk distribution preparation using leaderboard main and two ERC‑20s from references/workflows.md.
  • Build curl skeletons for store-distribution given mined hashes.
  • Explain why baseToken in storage payloads differs from empireAddress.

Error handling

HTTP Meaning
400 Bad parameters
401 API key / signature invalid
403 Forbidden
404 Not found
429 Rate limited
500 Server error — retry with backoff

Detail sections — open references next

Usage Guidance
Install only if you intentionally want an agent to help operate Empire Builder on live Base/Arbitrum mainnet. Do not provide private keys to the agent, keep API keys out of URLs when possible, require manual wallet confirmation for every transaction, and verify all recipients, amounts, chains, vault addresses, and signed-message text before approving.
Capability Tags
cryptorequires-walletcan-make-purchasescan-sign-transactionsrequires-sensitive-credentials
Capability Assessment
Purpose & Capability
The live crypto treasury capabilities match the stated purpose, but they are high-impact: SKILL.md says writes are live on Base/Arbitrum mainnet and use vault owner authority.
Instruction Scope
The workflow includes owner-signed executeBatch transactions and admin API writes; references/http-api.md also notes some signature messages are not server-template or nonce enforced, so operator-side scoping is critical.
Install Mechanism
There is no automatic install spec or code, but SKILL.md documents optional pinned npm dependencies for Clanker/viem workflows and recommends a lockfile/trusted diff.
Credentials
Production-only mainnet use is explicitly disclosed and purpose-aligned, but there is no sandbox/testnet path for safe dry runs beyond read-only GETs.
Persistence & Privilege
API keys plus guardian or vault-owner signatures can mutate backend records, leaderboards, staking state, burns, airdrops, and on-chain treasury balances; some credential/signature handling is left to user discipline.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install empire-builder
  3. After installation, invoke the skill by name or use /empire-builder
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.3.0
**Empire Builder 1.3.0** — Documentation scope clarified and safety section refactored. - Shrinks and simplifies SKILL.md by refactoring/removing long security and publisher note sections; security recommendations and registry flags are now assumed to be well-known product behaviors instead of “documentation bugs.” - Adds explicit "When NOT to use this skill" section listing inappropriate use cases (web app flows, co-signers, testnets, etc). - Expands and clarifies guidance on Empire ID vs SmartVault address, including a table of formats and places to find them. - Tightens language on agent constraints, mandatory flow steps, and proper API/auth usage. - Reference files and integration logic unchanged; still references only the exact `references/*` documentation for workflows and HTTP API.
v1.2.0
## Empire Builder v1.2.0 Changelog - Adds a new "Publisher note" section addressing security scanner and registry findings (e.g., ClawScan), explaining risk flags and how the publisher expects operators to mitigate them. - Moves and updates recommendations for human review and simulation before transaction signing, making human approval "recommended" for all treasury-moving operations. - Clarifies guidance around mainnet-only operation and absence of any sandbox/testnet, reiterates "production only" usage in several places. - Further documents expected high-risk findings for security registries, with a summary mitigation table (ASI02, ASI03, ASI04, ASI08). - Refines and strengthens language around supply chain safety, dependency pinning, and local vs. remote skill copy review. - No code or API workflow changes; this is a documentation and safety guidance update.
v1.0.2
**Added security guidance and pinning for dependencies in preparation for production agent workflows.** - Introduced a new section detailing security guardrails and recommended operational practices for transaction safety, human review, key management, and dependency supply chain hygiene. - Provided explicit instructions to pin dependency versions (`[email protected]` and `[email protected]`) and avoid `@latest` for any automation or agent-driven install. - Clarified guidance to review and diff skill or dependency updates before adopting changes, reinforcing a review-before-refresh stance for supply chain security. - Emphasized the risks and required practices for owner credential management and API key usage in treasury operations. - General structure, API routes, and integration described previously remain unchanged.
v1.0.1
No user-facing or functional changes in this release. - Version bump to 1.0.1 with no modifications to files or documentation. - All skill features, usage instructions, and agent guardrails remain unchanged.
v1.0.0
Empire Builder Skill 1.0.0 – Initial Release - Introduces the Empire Builder skill for managing SmartVault treasuries, leaderboards, boosters, distributions, burns, airdrops, and Clanker deploys. - Defines strict integration patterns and guardrails for agent use, including clear separation between Empire ID and SmartVault address. - Provides comprehensive documentation structure with references to specific workflow and HTTP API behaviors. - Limited to production endpoints on Base and Arbitrum mainnets; testnets and sandbox APIs are not supported. - Stresses exact adherence to provided HTTP API and workflow documentation to avoid common integration mistakes.
Metadata
Slug empire-builder
Version 1.3.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 5
Frequently Asked Questions

What is Empire Builder?

Empire Builder AI skill — SmartVault treasuries, leaderboards, boosters, prepare→executeBatch→store distributions, burns, airdrops, Clanker deploys. Self-con... It is an AI Agent Skill for Claude Code / OpenClaw, with 143 downloads so far.

How do I install Empire Builder?

Run "/install empire-builder" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Empire Builder free?

Yes, Empire Builder is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Empire Builder support?

Empire Builder is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Empire Builder?

It is built and maintained by CryptoRabble (@cryptorabble); the current version is v1.3.0.

💬 Comments