selat-openclaw-skills

OpenClaw / ClawHub skills powered by SELAT — domain skills (deal sourcing, lead gen, social intelligence, …) that pay for the data they use per call, in USDC, from the user's own self-custody Circle Agent Wallet. No API keys, no signups, no platform balance.

Each skill here is a thin intent wrapper over a vetted SELAT skill. The wrapper teaches an OpenClaw agent when to use the capability and how to run it safely; the machine-readable payment recipe (endpoints, rails, price caps) lives in SELAT-AI/selat-skills and is fetched by the selat CLI at install time. Wrappers contain no executable code and no endpoint URLs — installing one only ever writes text.

Skills

Skill Intent
vc-ai-infra-scout VC deal sourcing across an AI-infra / crypto-AI / robotics / agentic-payments thesis — discover founders, read raise chatter, distill lead-fund theses, enrich the top lead
twitter-research Read-only Twitter/X research — profiles, recent tweets, mentions, followers, tweet details/replies/retweeters, topic search, and trends. Curated 9-read menu; the agent runs only what's needed
perplexity-search Grounded web search & research via Perplexity — cheap web search plus agent-answer & async deep-research escalations, keyless & cited. Shown prices include SELAT's ~5% routing markup
selat-capabilities-for-metamask SELAT's capability layer for agents on a MetaMask Agent Wallet — tool use beyond the agent's native abilities, strongest as decision support for trading (perp funding rates, prediction-market candles, tokenized-stock quotes/technicals, news & social sentiment, macro regime — research data only, never order execution): discover paid APIs by intent, buy on rails the native x402 payer can't sign, fund a Gateway budget gaslessly, report spend. Unlike the wrappers above it needs no selat CLI or SELAT plugin: it runs on the mm CLI + selat-metamask-skills
stock-direction-signals Directional research brief on a US stock for agents on a MetaMask Agent Wallet — nine paid reads (quote, chart, RSI/MACD, news, earnings, Twitter/X, Reddit, macro) fused into a bullish/bearish/mixed brief. Research only, never orders or advice. Same mm-CLI engine — no selat CLI or SELAT plugin

More coming: lead enrichment, influencer discovery, social intelligence.

How every skill works — dry run before wallets

All skills in this repo follow the same three-step spine, in this order:

  1. Dry run first (free). selat skill verify probes every endpoint's live HTTP 402 challenge and prints the real quoted price per step — no wallet, no funds, no account. The agent shows the user actual prices before anything money-shaped exists.
  2. Wallet setup — only after the user opts in. selat init creates a self-custody Circle Agent Wallet (MPC — SELAT never holds keys or funds); selat fund deposits USDC into Circle Gateway.
  3. Paid run. selat skill run <name> executes the vetted recipe. The underlying skill defines each step's spend limit and the runner enforces it, printing a per-rail receipt summary. The live 402 quote is always the price source of truth.

Payments settle across multiple rails — direct Circle x402, routed MPP, and routed x402 on Base — auto-detected per step by the CLI.

Using a skill

Install from ClawHub with the OpenClaw skills commands, or point your agent at a SKILL.md in this repo. The only dependency is the selat CLI, one npm package:

npm install -g @selat-ai/selat-cli

Contributing a wrapper

Wrappers are thin, text-only SKILL.md files over a vetted SELAT skill. See AUTHORING.md for the SOP — frontmatter schema, required sections, the dry-run-first spine, the no-code/no-endpoint-URL rules, and the pre-PR checklist.

License

MIT — see LICENSE. Skills published to ClawHub are additionally available under MIT-0 per ClawHub's terms.