iiRecord
Agentic AI Atlas · Gas Town
methodology:gastowna5c.ai
II.
Methodology JSON

methodology:gastown

Structured · live

Gas Town json

Inspect the normalized record payload exactly as the atlas UI reads it.

File · domain/methodologies/methodologies.yamlCluster · domain
Record JSON
{
  "id": "methodology:gastown",
  "_kind": "Methodology",
  "_file": "domain/methodologies/methodologies.yaml",
  "_cluster": "domain",
  "attributes": {
    "displayName": "Gas Town",
    "description": "A multi-agent orchestration framework for AI-driven software development providing\ninfrastructure roles (Mayor, Deacon, Witness, Refinery), worker roles (Crew,\nPolecats, Dogs), and a git-backed work unit system (Beads, Convoys, Molecules).\nGoverned by GUPP (Gas Town Universal Propulsion Principle), MEOW (Molecular\nExpression of Work), and NDI (Nondeterministic Idempotence) for coordinating\nparallel agent execution with persistent attribution.\n",
    "methodologyKind": "orchestration",
    "origin": "Steve Yegge"
  },
  "outgoingEdges": [
    {
      "from": "methodology:gastown",
      "to": "domain:software-engineering",
      "kind": "applies_to"
    }
  ],
  "incomingEdges": [
    {
      "from": "lib-process:shared--gastown-convoy",
      "to": "methodology:gastown",
      "kind": "follows_methodology",
      "attributes": {
        "weight": 1
      }
    },
    {
      "from": "lib-process:shared--gastown-merge-queue",
      "to": "methodology:gastown",
      "kind": "follows_methodology",
      "attributes": {
        "weight": 1
      }
    },
    {
      "from": "lib-process:shared--gastown-molecule",
      "to": "methodology:gastown",
      "kind": "follows_methodology",
      "attributes": {
        "weight": 1
      }
    },
    {
      "from": "lib-process:shared--gastown-orchestrator",
      "to": "methodology:gastown",
      "kind": "follows_methodology",
      "attributes": {
        "weight": 1
      }
    },
    {
      "from": "lib-process:shared--gastown-patrol",
      "to": "methodology:gastown",
      "kind": "follows_methodology",
      "attributes": {
        "weight": 1
      }
    }
  ]
}