I Built a Village of Six Agents You Can Actually Inspect

Description and use

The project demonstrates a fully inspectable multi‑agent workflow where task ownership, memory, and spending can be audited without external AI calls.

Problem

How to transparently assign and inspect task ownership, memory, and spending in a multi‑agent simulation without relying on external model APIs.

Outcome

The project demonstrates a fully inspectable multi‑agent workflow where task ownership, memory, and spending can be audited without external AI calls.

Workflow

• 1. Choose an objective (e.g., Prepare for a raid).
• 2. Planner splits objective into jobs.
• 3. Residents bid on jobs using role suitability, distance, and experience.
• 4. Each job is assigned to one resident.
• 5. Residents execute tasks, gather resources, and coordinate recruitment.
• 6. If a resident rests, its job becomes available for reassignment.
• 7. Inspector displays current jobs, blocked work, spending limits, and recent memories.
• 8. Verify that job reassignment respects objective budget and constraints.

Evidence

The planner splits it into jobs. Six residents bid using role suitability, distance and bounded experience. Each job has one owner.

Source

The interesting part is the inspector: current jobs, blocked work, spending limits and recent memories are available to read alongside the action log.

Source

The default campaign uses deterministic game AI. It does not understand arbitrary natural-language goals, train a neural network or call a model behind the scenes.

Source

Use Node.js 24+: git clone https://github.com/Lingikaushikreddy/Settlement-Village cd Settlement-Village npm ci npm run dev

Source