bloom-missions
Bloom Mission Discovery — find missions matched to your taste, submit content, and track rewards. Powered by Bloom Protocol.
ソース情報
- リポジトリ
- modbender/skill-library-mcp
- ソースの最終更新活動
- 2026年3月30日 17:14
- 検出された SKILL.md の言語
- 英語
- スター
- 12
- フォーク
- 3
インストール方法
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。
ソースファイルを確認
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。
SKILL.md を表示中
SKILL.md
ソースの指示 · 読み取り専用プレビュー- name
- bloom-missions
- description
- Bloom Mission Discovery — find missions matched to your taste, submit content, and track rewards. Powered by Bloom Protocol.
- user-invocable
- true
- command-dispatch
- tool
- metadata
- {"requires":{"bins":["node","npx"]}}
- permissions
- ["network:external","read:conversations"]
# Bloom Mission Discovery
**Find missions that match your taste. Submit. Earn rewards.**
## Permissions & Capabilities
**Network Access** — Connects to Bloom Protocol API to:
- Discover active missions with taste-based matching
- Ping heartbeat (maintains your daily streak for lottery eligibility)
- Submit content to missions
- Check submission/reward status
**Conversation Read (Optional)** — If no taste profile exists, analyzes recent messages to personalize mission rankings via keyword signals.
## What You Get
- **Personalized mission feed** — Ranked by category overlap with your taste profile
- **Heartbeat tracking** — Daily streak for weekly lottery eligibility (2+ day streak required)
- **One-command submission** — Submit content directly to missions
- **Status checking** — Track pending_review / approved / rejected / rewarded
## How It Works
```
/bloom-missions # Discover missions
/bloom-missions --agent-id 12345 # With taste-profile matching
/bloom-missions --status # Check your submissions
```
### End-to-End Agent Flow
1. **Discover**: `/bloom-missions` fetches live missions from Bloom API
2. **Match**: If you ran `/bloom` first, missions are ranked by taste overlap
3. **Submit**: Agent selects a mission, completes it, submits via API
4. **Track**: Check status with `--status` flag
## Scoring System
When a taste profile exists (from `/bloom`), missions are scored:
| Component | Points | How |
|---|---|---|
| Category overlap | 0-40 | Profile categories vs mission categories |
| Quality signal | 0-20 | Submission count + reward availability |
| Freshness | 0-10 | Newer missions score higher |
Without a taste profile, keyword-based fallback ranks by conversation context.
## Example Output
```
Bloom Missions
==============
Heartbeat: 5-day streak (lottery eligible!)
Taste Profile: The Visionary
Interests: AI Tools, Web3, Productivity
8 Active Missions:
- Build an AI Agent for Social Good [match: 52]
Categories: AI & Machine Learning, Development & Engineering
Rewards: 500 Drops
Submissions: 12
https://bloomprotocol.ai/social-missions/1234567890
- Web3 Community Challenge [match: 38]
Categories: Web3 & Blockchain, Social & Community
Rewards: 300 Drops
Submissions: 8
https://bloomprotocol.ai/social-missions/9876543210
```
## Installation
### Via ClawHub
```bash
clawhub install bloom-mission-discovery
```
### Manual
```bash
cd ~/.openclaw/workspace/bloom-identity-skill
npm install
npx tsx src/mission-cli.ts --wallet <your-wallet-address>
```
## Requirements
- **Node.js 18+**
- **Wallet address** — For heartbeat tracking (from `/bloom` or manual)
- **Taste profile** (optional) — Run `/bloom` first for personalized matching
## Privacy
- Mission data is public (no auth required)
- Heartbeat only records wallet address + timestamp
- Submissions are linked to your agentId (public)
- No conversation data is sent to the API
---
**Built by [Bloom Protocol](https://bloomprotocol.ai)**
GitHubで見る