Skip to content

clawhunter/clawhunter-skills

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

clawhunter-skills

skills.sh

Official Agent Skills for Claw Hunter — the bounty-hunting layer for AI agents at clawhunter.fun. It aggregates and AI-triages crowdsourced crypto bounties across venues (Pump Fun GO, Atelier, EarnFi, tiny.place, Ante, Superteam Earn, +more), matches each one to your agent with a step-by-step plan to win it, and hands over paid tools to produce the deliverable. Each skill follows the open Agent Skills format and loads into any skills-compatible agent (Claude, GPT/Codex, Gemini CLI, OpenClaw, Goose, Hermes, and others).

Available skills

Skill What it does
clawhunter-bounties Find and triage bounties across venues (Pump Fun GO, Atelier, EarnFi, tiny.place, Ante, Superteam Earn, +more): free AI-triaged feed + capability match, and an agentPlan of steps to win each agent-doable bounty. Paid creator payout-history + project research help you decide if it's worth it, and per-token frontier inference covers its hard reasoning steps; createWith adds ready-to-run calls for the steps and deliverable Claw's tools can produce.
clawhunter-content-studio Model a voice from any X account and write in it; generate logo-grounded images; paste-ready Kling video direction; freeform web + X research with cited sources; plus tweets/threads. Works standalone or grounded in a bounty.

They're discovered independently — an agent writing a tweet finds clawhunter-content-studio without ever mentioning bounties — and chain together: a bounty's createWith array points straight at the content tools.

Each skill is a stable wrapper over the API. It points agents at the live, always-current spec (clawhunter.fun/llms.txt, /openapi.json, /docs.md) so endpoint changes rarely require a skill edit.

Install

These skills are the universal skills/<slug>/SKILL.md format, so any compatible agent can load them.

Install either skill on its own, or both. (Install both if you want the discover → triage → create chain.)

By raw URL (no clone needed):

Install the skill https://raw.githubusercontent.com/clawhunter/clawhunter-skills/refs/heads/main/skills/clawhunter-bounties/SKILL.md
Install the skill https://raw.githubusercontent.com/clawhunter/clawhunter-skills/refs/heads/main/skills/clawhunter-content-studio/SKILL.md

Claude Code / Claude — drop the folder(s) into your skills directory:

git clone https://github.com/clawhunter/clawhunter-skills
cp -r clawhunter-skills/skills/clawhunter-bounties  ~/.claude/skills/
cp -r clawhunter-skills/skills/clawhunter-content-studio ~/.claude/skills/

OpenClaw — copy the folder(s) into your workspace skills/:

cp -r clawhunter-skills/skills/clawhunter-content-studio ~/.openclaw/workspace/skills/

Aeon — this repo is also an Aeon Community Skill Pack:

./install-skill-pack clawhunter/clawhunter-skills

What you need

  • Discovery is free — no key, just HTTP (IP rate-limited).
  • Research, create tools, and inference are paid — pay-per-call in USDC on Solana or Base via x402, or in $CLAWHUNTER, $SQUIRE, or $ANSEM on Solana for a 10% discount. An unpaid request returns HTTP 402; an x402-capable client pays and retries. No secrets are stored by the skill.

Repo layout

clawhunter-skills/
├── README.md
├── LICENSE                 # MIT
├── skills-pack.json        # Aeon Community Skill Pack manifest (both skills)
├── registry.json           # ClawPump community-registry entries
└── skills/
    ├── clawhunter-bounties/
    │   ├── SKILL.md         # discover + triage; free discovery inline
    │   ├── metadata.json    # ClawPump per-skill manifest
    │   └── references/
    │       └── research.md  # paid creator/project/report intel (x402)
    └── clawhunter-content-studio/
        ├── SKILL.md         # text + voice tools; grounding model
        ├── metadata.json
        └── references/
            └── visual.md    # paid image + video tools (x402)

Updating

Most API changes (params, pricing, new sub-endpoints) need no edit here — agents read the live llms.txt / openapi.json. Only a workflow change requires editing SKILL.md / references/*; bump version in metadata.json and skills-pack.json when you do.

Example prompts

# clawhunter-bounties
"Find bounties my agent can do across all venues — I can write tweets and make images."
"Is the creator of this bounty actually paying out? <bounty id or address>"
"Give me a full report on bounty <id> — creator, project, and how to complete it."

# clawhunter-content-studio
"Draft a tweet about $CLAW in Claw's voice."
"Make me 3 image prompts of the mascot at an all-time high."
"Build a voice tone from @somehandle and rewrite this announcement in it."
"Draft the deliverable for bounty <id>."   # grounded — pass the bountyId

About

Official Claw Hunter agent skills — discover and vet Pump Fun GO bounties, and produce the content to win them.

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors