メニュー
Aggregate, score, and prioritize social mentions across X/Twitter. Outputs a ranked inbox with engagement scores and draft context for efficient reply batching.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/0xAxiom/axiom-public --skill social-inboxコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Claim creator LP fees for Net Protocol / Banger memecoins on Robinhood Chain from locked Uniswap V3 position NFTs, then optionally use the claimed WETH to buy back the memecoin. Use when the user says "claim Robinhood fees", "claim Hoodie fees", "claim Netr/Banger LP fees", links a Robinhood Blockscout UNI-V3-POS NFT instance, or asks to buy back a Robinhood memecoin with claimed fees. Do NOT use for Base Clanker fee claims or Uniswap V4 positions.
Launch Net Protocol Netr memecoins on Robinhood Chain. Use when the user says "launch a Robinhood token", "deploy a memecoin through Net Protocol", "create a Netr token on Robinhood", "buy the token after launch", or needs Bankr-backed or private-key-backed deployment to chain 4663. Supports direct NET_PRIVATE_KEY signing, Bankr wallet API signing, encode-only transaction generation, gas bridging through Bankr, and post-launch verification.
Encode EVM function calls and constructor args from signature + values. Zero deps.
SOC 職業分類に基づく
SKILL.md を表示中
| name | social-inbox |
| description | Aggregate, score, and prioritize social mentions across X/Twitter. Outputs a ranked inbox with engagement scores and draft context for efficient reply batching. |
Fetches X/Twitter mentions and brand searches, deduplicates against already-replied tweets, scores by engagement potential, and outputs a prioritized inbox.
# Full scan: mentions + brand search
node ~/Github/axiom-public/agent-skills/skills/social-inbox/social-inbox.mjs
# Mentions only
node ~/Github/axiom-public/agent-skills/skills/social-inbox/social-inbox.mjs --mentions-only
# Output as JSON (for piping)
node ~/Github/axiom-public/agent-skills/skills/social-inbox/social-inbox.mjs --json
# Custom search terms
node ~/Github/axiom-public/agent-skills/skills/social-inbox/social-inbox.mjs --search "AppFactory,axiom ventures"
Saves to ~/clawd/data/social-inbox.json with structure:
{
"lastScan": "2026-02-08T...",
"items": [
{
"id": "tweet_id",
"author": "@handle",
"text": "...",
"url": "https://x.com/...",
"score": 8.5,
"scoreBreakdown": { "relevance": 3, "question": 2, "directMention": 2, "recency": 1.5 },
"status": "pending",
"source": "mentions"
}
]
}
| Signal | Points |
|---|---|
| Direct @mention | +2 |
| Question mark (asking us something) | +2 |
| Keywords: fund, invest, agent, build, LP, token | +1 each (max 3) |
| Recency (<1h: +2, <6h: +1.5, <24h: +1) | up to +2 |
| Not a bot/spam pattern | +1 (default, -1 if detected) |
Items scoring <3 are filtered out. Already-replied tweets are excluded.
Add to OpenClaw cron for periodic scanning:
{
"kind": "agentTurn",
"message": "Run social inbox scan: node ~/Github/axiom-public/agent-skills/skills/social-inbox/social-inbox.mjs\nReview the top 3 items and draft replies if valuable."
}
twitter-api.py (already installed)twitter-replied.json dedup log (already exists)