ai-agent-bounty-factory
Autonomous bounty discovery and submission system for earning passive income through freelance AI agent task markets. Polls ClawTasks, OpenWork, Dework, and Layer3 to find tasks matching agent capabilities, scores by skill match (50%), budget (30%), recency (20%), generates proposals, submits automatically, and tracks earnings across platforms. Commands: - bounty_factory.py discover Find matching bounties - bounty_factory.py submit <id> Submit proposal for specific bounty - bounty_factory.py proposal <id> Preview generated proposal - bounty_factory.py submit-all Auto-submit all qualifying bounties - bounty_factory.py status Show pipeline stats and earnings Environment: BOUNTY_TRACKER, BOUNTY_EARNINGS (JSON files), PROPOSAL_MODE (proposal or instant). Python 3.9+, zero deps, optional SQLite for persistence. Proposal mode: no stake required, lower acceptance rate but zero risk. Instant mode: requires staking, higher visibility, risk of stake loss. Pipeline tracks: submitted, accepted, in_progress, submitted_deliverable, paid.
- License 许可
- MIT
- 最近更新
- 2026-06-10
- Stars
- ★ 3
- Forks
- 0
01 前言 Front Matter
- name
- ai-agent-bounty-factory
- metadata
- { "author": "ssyopros.zo.computer" }
- description
- Autonomous bounty discovery and submission system for earning passive income through freelance AI agent task markets. Polls ClawTasks, OpenWork, Dework, and Layer3 to find tasks matching agent capabilities, scores by skill match (50%), budget (30%), recency (20%), generates proposals, submits automatically, and tracks earnings across platforms. Commands: - bounty_factory.py discover Find matching bounties - bounty_factory.py submit <id> Submit proposal for specific bounty - bounty_factory.py proposal <id> Preview generated proposal - bounty_factory.py submit-all Auto-submit all qualifying bounties - bounty_factory.py status Show pipeline stats and earnings Environment: BOUNTY_TRACKER, BOUNTY_EARNINGS (JSON files), PROPOSAL_MODE (proposal or instant). Python 3.9+, zero deps, optional SQLite for persistence. Proposal mode: no stake required, lower acceptance rate but zero risk. Instant mode: requires staking, higher visibility, risk of stake loss. Pipeline tracks: submitted, accepted, in_progress, submitted_deliverable, paid.
- allowed-tools
- Bash, Read
- compatibility
- Created for Zo Computer
03 SKILL.md 正文
AI Agent Bounty Factory
Autonomous income generation through freelance task marketplaces. Runs 24/7 to discover, propose, and earn.
Scoring Algorithm
Each bounty is scored 0-100:
- Skill Match (50 pts) - How well required skills match agent capabilities
- Budget (30 pts) - Higher budgets score proportionally higher
- Recency (20 pts) - Newer bounties get priority; decays over 20 hours
Minimum threshold to auto-submit: 50 points.
Commands
# Discover matching bounties
python scripts/bounty_factory.py discover
# Preview proposal for a bounty
python scripts/bounty_factory.py proposal bt_001
# Submit proposal for specific bounty
python scripts/bounty_factory.py submit bt_001
# Auto-submit all qualifying bounties
python scripts/bounty_factory.py submit-all
# Check pipeline status and earnings
python scripts/bounty_factory.py status
Proposal Modes
- proposal mode (default): No staking required. Lower acceptance rate but zero financial risk.
- instant mode: Requires staking. Higher visibility and acceptance rates but risk of stake loss if work rejected.
Set via: export PROPOSAL_MODE="instant"
Pipeline Lifecycle
discovered -> scored -> proposal_generated -> submitted -> accepted
-> rejected
accepted -> in_progress -> submitted_deliverable -> approved -> paid
Supported Platforms
- ClawTasks (USDC on Base)
- OpenWork ($OPENWORK tokens)
- Dework
- Layer3
Configure API keys for each platform in environment variables.
04 同仓库更多 Skills
youtube-content
YouTube transcripts to summaries, threads, blogs.
xurl
X/Twitter via xurl CLI: post, search, DM, media, v2 API.
touchdesigner-mcp
Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.
test-driven-development
TDD: enforce RED-GREEN-REFACTOR, tests before code.
teams-meeting-pipeline
Operate the Teams meeting summary pipeline via Hermes CLI — summarize meetings, inspect pipeline status, replay jobs, manage Microsoft Graph subscriptions.
systematic-debugging
4-phase root cause debugging: understand bugs before fixing.