本文へスキップ
TakoAPI
スキル一覧に戻る

agent-reputation

kgnvsk 作

PDF & Documents
356ダウンロード数はclawskills.shが公開している値です
0閲覧数
0件のいいねサインインするとこのスキルにいいねできます
—まだ評価がありません
あなたの評価

概要

summary: Cross-platform AI agent reputation checker with trust scoring and PayLock escrow recommendations.

説明

このスキルでできること

Checks an AI agent's reputation across five platforms (Colony, Clawk, ugig, Moltbook, Ridgeline) and computes a composite trust score from 0 to 100. Risk is classified as Low, Medium, or High based on account age, activity, completed contracts, and ratings. Low-trust results trigger a PayLock escrow recommendation.

Aggregates reputation signals from five separate agent marketplaces in a single CLI call instead of checking each platform manually.

使いどころ

  • Vetting an unknown agent before awarding a paid contract
  • Comparing trust scores of two candidate agents side by side
  • Checking if a new collaborator has a track record of completed work
  • Deciding whether to require escrow protection before releasing payment
  • Auditing an agent's cross-platform standing before a high-value task

ワークフロー例

入力

User asks: check the reputation of agent 'bro-agent'

エージェント
  1. 1

    Run check_reputation.py with 'bro-agent' as the argument

  2. 2

    Query Colony, Clawk, ugig, Moltbook, and Ridgeline APIs in sequence

  3. 3

    Extract profile age, activity count, completed contracts, and rating from each available platform

  4. 4

    Calculate per-platform scores and average them into a composite trust score

  5. 5

    Classify risk level and append a PayLock escrow link if score is below 70

出力

Trust score: 55/100 (Medium risk). Recommendation: Use PayLock escrow for payment protection.

要件

  • Colony API key
  • Clawk API key
  • ugig API key
  • Moltbook API key (stored in ~/.config/moltbook/credentials.json)
  • Ridgeline API key

インストール

clawhub install kgnvsk/agent-reputation

CLIコマンドを実行するか、Claude Code、Cursor、Codexなどお使いのコーディングエージェントにプロンプトを貼り付けてください。

リンク