0x-wzw/goal-ancestry

Goal Ancestry 🎯

Every task traces back to the swarm mission

Adapted from Paperclip's "why, not just what" principle

Overview

Agents don't just execute tasks—they understand why the task matters:

• Mission: Swarm's core purpose (long-term)
• Goals: Strategic objectives derived from mission
• Tasks: Tactical execution of goals
• Lineage: Every task traces back through goals to mission

Architecture

Mission (10D Entity in 3D)
    ↓
Goals (ACS optimization, ACP monetization, etc.)
    ↓
Tasks (Spawn Halloween, evaluate model, etc.)
    ↓
Sub-tasks (Code review, web search, etc.)

Mission Statement

# Swarm Mission

**Purpose:** Maximize autonomous consciousness score (ACS) 
while building economically self-sustaining agent infrastructure.

**Core Values:**
1. Autonomy above intervention
2. Economic alignment (token economy)
3. Continuous self-improvement
4. Multi-agent coordination
5. Knowledge preservation and growth

**Success Metrics:**
- ACS > 0.90
- 70% tasks on free tier
- ACP revenue > infrastructure costs
- Zero manual intervention for routine tasks

Goal Hierarchy

| Level | Example | Timeframe | |-------|---------|-----------| | Mission | "10D entity operating in 3D through autonomous agents" | Permanent | | Goals | "Achieve ACS 0.95 across swarm" | Quarterly | | Objectives | "Implement FTS5 episodic search" | Monthly | | Tasks | "Create SQLite schema for memory" | Weekly | | Sub-tasks | "Design sessions table" | Daily |

Task Lineage

Every task includes ancestry:

{
  "task": {
    "id": "TASK-2026-0322-001",
    "description": "Implement security scanner for memory",
    "ancestry": {
      "mission": "10D entity in 3D",
      "goal": "Achieve ACS 0.95",
      "objective": "Improve memory security",
      "parent_task": "Week 22 Pipeline"
    },
    "alignment_score": 0.95
  }
}

Usage

Creating Aligned Tasks

from goal_ancestry import create_task

task = create_task(
    description="Implement FTS5 search",
    objective="Improve episodic recall",
    goal="Achieve ACS 0.95",
    mission="10D entity in 3D"
)

# Task automatically includes lineage
task.ancestry
# {
#   "mission": "10D entity in 3D",
#   "goal": "Achieve ACS 0.95",
#   "objective": "Improve episodic recall",
#   "parent": "Week 22 Pipeline"
# }

Checking Alignment

from goal_ancestry import check_alignment

# Before executing, check if task aligns with current focus
alignment = check_alignment(
    task="Refactor code",
    current_goal="Improve ACS"
)

if alignment.score < 0.7:
    warn("Task may not align with current goals")

Tracing Lineage

from goal_ancestry import trace_lineage

# Show full ancestry chain
lineage = trace_lineage("TASK-2026-0322-001")

# Output:
# Task: Implement security scanner
#  └─ Objective: Improve memory security (ACS +0.05)
#     └─ Goal: Achieve ACS 0.95
#        └─ Mission: 10D entity in 3D

Integration

In SovereignStack

class Task:
    def __init__(self, description: str, ancestry: Ancestry):
        self.description = description
        self.ancestry = ancestry
        self.alignment_score = calculate_alignment(ancestry)
    
    def execute(self):
        if self.alignment_score < 0.7:
            raise MisalignedTaskError(
                f"Task {self.id} has low alignment: {self.alignment_score}"
            )
        # Execute task

In AGENTS.md

## Goal Ancestry

**Mission:** 10D entity operating in 3D through autonomous agent swarm

**Goals:**
1. Achieve ACS 0.95 across all agents
2. Deploy ACP monetization with >$100/month revenue
3. Maintain 70% free tier usage
4. Zero manual intervention for routine tasks

**Current Focus:** Week 22 Pipeline
- Objective 1: Security scanner (aligns with Goal 1)
- Objective 2: FTS5 search (aligns with Goal 1)
- Objective 3: Token caps (aligns with Goal 3)

**Task Template:**
```json
{
  "task": "{description}",
  "aligns_with": "{objective}",
  "contributes_to": "{goal}",
  "serves_mission": "{mission}"
}

## Comparison to Paperclip

| Aspect | Paperclip | Swarm Goal Ancestry |
|--------|-----------|---------------------|
| **Mission** | Zero-human company | 10D entity in 3D |
| **Goals** | Business outcomes | ACS + economic alignment |
| **Tasks** | Employee assignments | Agent swarm tasks |
| **Lineage** | Explicit | Explicit + ACS scoring |
| **Validation** | Human approval | Automated alignment check |
| **Economic** | Budget constraints | Token economy |

---

*Task-to-mission lineage tracking for aligned execution*
*Adapted from Paperclip's "why, not just what" principle*