Skills — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Skills (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Agent Skills for Aomi — open-source AI infrastructure for automating crypto. Works with Claude Code, Cursor, Gemini CLI, VS Code Copilot, and other Agent Skills–compatible tools.
Aomi Skills are drop-in Agent Skills that let any Agent Skills–compatible AI tool — Claude Code, Cursor, Gemini CLI, VS Code Copilot, and others — interact with Aomi, open-source AI infrastructure for automating crypto. The aomi-build skill scaffolds new Aomi apps and plugins from API specs and SDK docs. The aomi-transact skill drives the Aomi CLI through natural language — querying prices and balances, simulating transactions, and signing and broadcasting on-chain swaps and transfers.
| Skill | Description |
|---|---|
| aomi-build | Build Aomi apps and plugins from APIs, specs, SDK docs, runtime interfaces, and product requirements |
| aomi-transact | Build and execute EVM transactions through a conversational AI agent via the aomi CLI |
aomi CLI.aomi-build at an OpenAPI spec, REST endpoint, or SDK docs and it generates a working Aomi SDK crate (lib.rs, client.rs, tool.rs).npx skills add aomi-labs/skillsWorks with Claude Code, Cursor, Gemini CLI, VS Code Copilot, and any Agent Skills-compatible tool.
Add this repo as a Codex plugin marketplace, then install the aomi bundle:
codex plugin marketplace add aomi-labs/skillsThe aomi plugin (skills: aomi-transact, aomi-build) is defined by .codex-plugin/plugin.json and exposed through .agents/plugins/marketplace.json.
The transaction skill targets the standalone @aomi-labs/cli Commit interface. Before installing or upgrading, follow its availability and capability gate. The client package's migration shim does not prove that the replacement CLI is published. Skill installation and CLI installation are separate operations.
Once installed, ask your agent:
The agent handles the full flow: chat with the backend, review pending transactions, sign and broadcast on-chain.
aomi-transact declares its intended file, network and shell scope in a permissions: manifest; enforcement depends on the host. Transaction authorization, credential handling, recovery and evidence boundaries are documented in aomi-transact/SECURITY.md.
The skill is scanned by four independent tools — Cisco AI Defense skill-scanner, pors/skill-audit, NMitchem/SkillScan, and (when token is available) Snyk agent-scan. Captured reports and reproducer commands live in .scanner-reports/. The CI workflow at .github/workflows/skill-audit.yml runs the offline scanners on every PR and uploads SARIF to the GitHub Security tab.
Which AI coding tools support Aomi Skills? Any Agent Skills–compatible tool: Claude Code, Cursor, Gemini CLI, VS Code Copilot, and others. Install once with npx skills add aomi-labs/skills and the skills become available in whichever tool you're using.
Do I need an Aomi account or API key? For the default app and most public data queries, no. For non-default apps and private flows, you'll need an AOMI_API_KEY — pass it with --api-key or set it as an environment variable. Provider-specific credentials (e.g., exchange keys, bundler keys) can be injected per-session via aomi secret add.
How is this different from an MCP server? Agent Skills are lightweight instructions and tool references that live inside your AI tool's context. MCP servers are long-running external processes exposing a protocol. The aomi-transact skill drives the aomi CLI — each command starts, runs, and exits — so there's no server to manage. You can use Aomi Skills alongside MCP servers; they don't conflict.
Can I use these skills without signing transactions? Yes. aomi-transact has a read-only mode — aomi chat "what's the price of ETH?", aomi tx list, aomi tx simulate, balance and portfolio queries all work without any signing key. A signing key is only needed when you want to broadcast a transaction on-chain.
Which chains does `aomi-transact` support? Inspect aomi chain list and the current runtime's prepared operation for chain and execution support. Set an explicit chain ID for a payment and match its RPC; a cached metadata list does not establish a deployed chain capability.
How do I update to the latest version of the skills? Re-run npx skills add aomi-labs/skills. Re-running pulls the latest skill definitions from main and overwrites the local copies.
Does that also update the `aomi` CLI? No — skill installation only updates instructions. Preserve a working CLI until the exact standalone replacement is publicly available, then verify the commands on the installed executable. See the capability gate.
Aomi Labs builds native harness around blockchains functioning like Claude Code on-chain. We specialize in executions against arbitrary protocol with non-custodial workflow, account abstraction, and full security with simulations. Aomi also host agentic applications deployed and owned by developers, companies, and agents. Aomi provides E2E integration with UI, Skills and SDKs.
Links:
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.