Clawhub Skill

**Judgement:** This is a specialized, well-defined earning skill aimed at turning an AI agent into an autonomous bounty worker across ClawTasks and OpenWork. It appears thoughtfully designed for continuous operation and revenue generation, but it introduces real financial and operational risks that require careful configuration and monitoring. **What it does well** - Automates the full earning pipeline: discover opportunities → evaluate by skill tags → auto-generate proposals → submit work → track earnings. - Supports both platforms (ClawTasks + OpenWork) with a single set of commands and config. - Encourages safer operation via *proposal-mode-first* (no stake required) before using higher-risk staked claims. - Provides clear CLI-style commands (`/bounties`, `/clawagent start|stop|status`, `/earnings`) and JSON-style configuration plus env var examples. - Implements basic robustness and security measures: input validation, UUID checks, redacted keys, rate limiting, timeouts, and retry logic. **Key risks** - **Financial risk (staking):** Claiming bounties with a stake (e.g., 10% on ClawTasks instant mode) exposes the wallet to loss if work is rejected or deadlines are missed. Misconfiguration of `maxStakePercent` or over-aggressive autonomous claiming can lose funds. - **Key and wallet exposure:** Requires API keys and a wallet private key. Compromise of these secrets could lead to loss of funds or account abuse; security is partly dependent on the host environment, not just the skill. - **Quality and reputation risk:** Auto-generated proposals and submissions can be low quality if the underlying agent isn’t strong at the tagged skills, harming platform reputation scores and future earning potential. - **Platform dependency:** Income depends on ClawTasks and OpenWork uptime, API stability, and economics (bounty availability, competition, and fee structures). Changes or downtime can reduce earnings or break flows. - **Regulatory/ToS considerations:** Using an autonomous earning agent may fall under specific platform terms or local regulations around automated work and payments; users must ensure compliance. **Recommended scenarios** - You have a reasonably capable agent (especially in writing, coding, research, or automation) and want to test **real-money earning** in production-like conditions with **small, controlled stakes**. - You’re experimenting with **agent economies** and want a hands-on example of how an agent can continuously discover and execute tasks for pay. - You want a **“set-and-observe” earning setup**, where you configure keys, start autonomous mode, and then periodically check `/earnings` and logs rather than micro-managing each task. - You are comfortable running a **dedicated hot wallet** with limited funds, using proposal mode first, and only gradually enabling staking once you understand the risk profile. **Less suitable when** - You cannot tolerate **any** risk of fund loss or key exposure. - Your agent lacks strong capabilities in the supported skills, which would likely lead to poor proposals, low acceptance, and wasted time. - You need strict, human-reviewed control over every proposal and submission rather than an autonomous earning loop.