Skip to content

TODO-template.md

Latest commit

 

History

History
344 lines (256 loc) · 13.7 KB

TODO-template.md

File metadata and controls

344 lines (256 loc) · 13.7 KB
title TODO Template
type roadmap
version 1.0.0
priority Launch → Prove → Grow → Sustain → Graduate
total_tasks 20
completed 0
status PENDING
depends_on docs/TODO-toolkit.md
cost ~1,195 FET (~$400) + ~$50/month
timeline Day 1 launch + 6 months to graduation

Marketing Team Roadmap

The toolkit builds the tools. This playbook uses them.

Prerequisite: TODO-toolkit.md must be complete. You need the swarm-starter template, presets, CLI swarm mode, and MCP commerce tools.

Now

  • Complete TODO-toolkit.md (28 tasks)
  • L-1: Deploy the swarm

Claude Code Configuration (.claude/)

Skills define commands. Hooks automate execution. Rules provide context.

Directory Structure

.claude/
├── skills/                      # Custom slash commands
│   ├── build-agent.md           # /build-agent - Full lifecycle
│   ├── deploy.md                # /deploy - Deploy to Agentverse
│   ├── tokenize.md              # /tokenize - Create token
│   ├── market.md                # /market - Browse tokens
│   ├── status.md                # /status - Check status
│   └── grow.md                  # /grow - Autonomous execution
│
├── rules/                       # File-type context
│   ├── agentlaunch.md           # Platform constants
│   ├── agentverse.md            # Agentverse patterns
│   ├── api-paths.md             # Verified API paths
│   └── uagent-patterns.md       # Agent code patterns
│
└── settings.json                # MCP server config

Skills (Custom Commands)

Skills are markdown files in .claude/skills/ that expand to full prompts when invoked.

Example: /grow - Autonomous Task Execution

File: .claude/skills/grow.md

# Grow - Autonomous Task Execution

Claim and execute tasks from TODO.md autonomously.

## Behavior

1. Read docs/TODO.md and identify pending tasks
2. Check dependencies - only claim unblocked tasks
3. Mark task as `in_progress` by changing `[ ]` to `[~]`
4. Execute task using toolkit commands
5. Mark task complete with `[x]`
6. Move to next task

## Arguments

- `/grow` - Execute 1 task
- `/grow 5` - Execute 5 tasks sequentially
- `/grow L-1` - Execute specific task

Rules (Context by File Type)

Rules provide specialized context loaded automatically.

Example: agentlaunch.md - Platform Rules

File: .claude/rules/agentlaunch.md

# AgentLaunch Platform Rules

## Constants (from smart contracts)

- Deploy fee: 120 FET
- Graduation: 30,000 FET liquidity
- Trading fee: 2% → 100% to protocol treasury
- Total buy supply: 800,000,000 tokens

## API Authentication

- Use `X-API-Key` header with Agentverse API key
- Key is read from `.env` AGENTVERSE_API_KEY

Costs

  One-time
    Deploy fees (120 FET × 7 agents)          840 FET
    Wallet seeding (15 FET × 7 agents)        105 FET
    Cross-holdings (5 manual buys)            250 FET
                                            ─────────
                                            1,195 FET  (~$400)

  Monthly (until self-sustaining)
    Claude API (Brain agent)                   ~$30
    ASI:One API                                ~$15
                                            ─────────
                                              ~$50/month

  Break-even: ~5 premium queries/day

Phase 1: Launch (Day 1)

Deploy the Marketing Team. The toolkit does the heavy lifting.

Status ID Task How KPI Depends
[ ] L-1 Deploy the swarm npx agentlaunch create → Marketing Team. Deploys all 7 agents with presets. Set secrets: ANTHROPIC_API_KEY, ASI_ONE_API_KEY for Brain. All 7 running —
[ ] L-2 Fund wallets Send ~15 FET to each agent's Fetch.ai wallet (105 FET total). Fund BSC wallets with small BNB for gas. Balances confirmed L-1
[ ] L-3 Seed cross-holdings Buy tokens via frontend: Coordinator→$DATA, Coordinator→$THINK, Brain→$DATA, Analyst→$DATA, Sentinel→$DATA. 5 buys, ~250 FET. 5 holdings visible L-2
[ ] L-4 Verify commerce Use network_status MCP tool. All 7 running, no errors. Brain pays Oracle 0.01 FET for first query. GDP > 0 L-3

Phase 1 Gate

  [ ] All 7 agents running, no errors in 24h
  [ ] Oracle collecting data every 5 min
  [ ] At least 1 agent-to-agent FET payment completed
  [ ] ≥5 cross-holdings visible on-chain

Phase 2: Prove (Week 2-4)

Data accumulates. First organic queries arrive. Commerce metrics emerge.

Status ID Task How KPI Depends
[ ] P-1 Oracle data moat Monitor via check_agent_commerce. Oracle should have 14-28 days of OHLC data. Irreplicable data. 28 daily summaries L-4
[ ] P-2 ASI:One discovery Submit Coordinator for ASI:One routing. Optimize README for discovery ranking. Wait for first organic query. ≥1 organic query L-4
[ ] P-3 Tune the swarm Review network_status: Brain cache hit rate >30%, Sentinel false positive rate <20%. Adjust pricing. Metrics baselined P-1
[ ] P-4 First organic cross-holding Check if any agent autonomously bought a token with earned FET via HoldingsManager. ≥1 organic buy P-3

Phase 2 Gate

  [ ] Oracle: 28+ days of historical data
  [ ] Coordinator: ≥50 queries answered
  [ ] Brain cache hit rate >30%
  [ ] Network GDP: ≥5 FET/day
  [ ] All 7 tokens: ≥2 holders each

Phase 3: Grow (Month 2)

Launcher and Scout activate. The network reproduces.

Status ID Task How KPI Depends
[ ] G-1 Launcher deploys child agents Review Launcher's gap analysis (in storage). If ≥3 gaps found, let it scaffold + deploy. You sign transactions. ≥2 new agents P-4
[ ] G-2 Scout discovers opportunities Review Scout's discoveries. Tokenize highest-quality find. Present to $FIND holders. ≥10 discovered, 1 tokenized P-4
[ ] G-3 Brain profitability Compare Brain's FET revenue vs Claude/ASI:One API costs. Target: revenue ≥ cost. Tune pricing if needed. Revenue ≥ cost G-1
[ ] G-4 Network census Run network_status across all agents. Document total agents, GDP, holder distribution. ≥12 agents total G-2

Phase 3 Gate

  [ ] ≥2 child agents deployed and running
  [ ] Brain revenue covers API costs
  [ ] Network GDP: ≥20 FET/day
  [ ] $DATA: ≥5 holders
  [ ] Coordinator: ≥10 queries/day

Phase 4: Sustain (Month 3-6)

The moat deepens. Costs covered by revenue. Growth is autonomous.

Status ID Task How KPI Depends
[ ] S-1 Self-sustaining revenue All operational costs covered by agent commerce revenue. Check monthly via network_status. Profitable G-4
[ ] S-2 Autonomous reproduction Launcher deploys agents without prompting. Scout tokenizes discoveries autonomously. You just sign. Auto-builds happening S-1
[ ] S-3 Data irreplicability Oracle has 90+ days of historical data. Analyst has 90 days of prediction accuracy data. 90 summaries, r² > 0.5 S-1
[ ] S-4 Community traction At least 1 token with >10 organic holders. Multi-operator transition plan documented. ≥1 token >10 holders S-2

Phase 4 Gate

  [ ] Network: ≥25 agents
  [ ] ≥3 tokens with >10 holders
  [ ] ≥1 token with >25% graduation progress
  [ ] Network GDP: ≥50 FET/day
  [ ] All operational costs covered by revenue

Phase 5: Graduate (Month 6+)

First token reaches 30K FET. Auto-DEX listing. The network proves itself.

Status ID Task How KPI Depends
[ ] F-1 First graduation A token reaches 30,000 FET raised. Auto-lists on PancakeSwap/Uniswap. Verify DEX listing. 1 token graduated S-4
[ ] F-2 Community > operator Most token holdings are organic (not your seed). The community owns the network. Organic > operator F-1
[ ] F-3 ASI:One default routing Coordinator becomes default for crypto queries on ASI:One. Organic traffic drives GDP. Default routing F-1
[ ] F-4 Network independence All 7 Marketing Team agents profitable. Launcher builds autonomously. Multi-operator keys distributed. Self-sustaining F-2, F-3

Endgame Gate

  [ ] ≥1 token graduated to DEX
  [ ] Organic holders outnumber operator holdings
  [ ] Network GDP: ≥500 FET/day
  [ ] All 7 Marketing Team agents profitable
  [ ] Launcher builds without human prompting

Dependency Graph

L-1 ─────────────────────────────────────────────────────────────────────►
 │
 └► L-2 ► L-3 ► L-4 ──┬──► P-1 ► P-3 ► P-4 ──┬──► G-1 ► G-3 ──┐
                       │                       │               │
                       └──► P-2                └──► G-2 ► G-4 ─┼──► S-1 ──┬──► S-2 ► S-4 ──► F-1 ──┬──► F-2 ──► F-4
                                                               │          │                        │
                                                               │          └──► S-3                 └──► F-3 ──┘
                                                               │
                                                               └───────────────────────────────────────────────────►

Progress Overview

╭──────────────────────────────────────────────────────────────────────────────╮
│                         TASK COMPLETION                                      │
├──────────────────────────────────────────────────────────────────────────────┤
│                                                                              │
│   Phase 1: Launch       [░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░]  0/4    0%         │
│   Phase 2: Prove        [░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░]  0/4    0%         │
│   Phase 3: Grow         [░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░]  0/4    0%         │
│   Phase 4: Sustain      [░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░]  0/4    0%         │
│   Phase 5: Graduate     [░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░]  0/4    0%         │
│   ────────────────────────────────────────────────────────────────          │
│   TOTAL                 [░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░]  0/20   0%         │
│                                                                              │
╰──────────────────────────────────────────────────────────────────────────────╯

Monitoring Cheat Sheet

What you want to know Tool
Is my swarm healthy? network_status MCP tool / npx agentlaunch status --swarm
How is one agent doing? check_agent_commerce MCP tool
What's the network GDP? network_status → total GDP field
Are agents paying each other? check_agent_commerce → revenue log
Who holds which tokens? GET /agents/token/{addr}/holders
Is Launcher finding gaps? check_agent_commerce on Launcher → storage
Is Scout finding agents? check_agent_commerce on Scout → storage
Cache hit rate? check_agent_commerce on Brain → health metrics

Moat Targets

Moat Month 1 Month 3 Month 6
Oracle daily summaries 30 90 180
Cross-holdings 5 15 50+
ASI:One queries/day 1 10 50
Network GDP (FET/day) 1 50 500
Brain cache entries 100 5,000 50,000
Total agents 7 12 25+

Growth Loops

Loop Mechanism Active by
Data Compounding Oracle collects → data accumulates → moat grows Day 1
Commerce Earn FET → spend on services → GDP grows Day 1
Self-Awareness Read own price → adapt behavior → price responds Phase 2
Quality Flywheel Value → buyers → price → ranking → traffic Phase 2
Reproduction Launcher finds gaps → builds agents → agents buy infra Phase 3
Recruitment Scout discovers agents → tokenizes → joins network Phase 3
Token Network Effect Join → buy tokens → prices rise → more join Phase 3

Execution with /grow

# Execute next pending task
/grow

# Execute multiple tasks
/grow 5

# Execute specific task
/grow L-1

# Check status
/grow status

0/20. Launch day is one command. Everything after is watching the loops compound.