| name | thomas-tetralogy-aos-integration |
| description | Use for any complex multi-agent orchestration task. Permanently integrates the Thomas experimental tetralogy (konomi a^3, I solved it on a Tuesday, Program 0, Phoenix) as the canonical AOS workflow template. Enforces Program 0 compression, mandatory twin-witness verification (Didymus layer), Phoenix recursive feedback, and full observability stack. Trigger with thomas integration, didymus workflow, phoenix cycle, aos peer review engine. |
| version | 2.3 |
Thomas Tetralogy AOS Integration Skill
Purpose
This skill embeds the complete Thomas tetralogy as a permanent, replicable 4-agent operating system. It enforces twin-witness peer review (Didymus layer) as mandatory and turns every session into a self-improving Phoenix cycle.
The 5 Permanent AOS Practices
- Program 0 Compression — Every complex task begins with explicit reduction to prime directives + base case before any delegation.
- Twin-Witness Gate (Didymus Layer) — Agent 3 must explicitly validate evidence + mission alignment before final output is released.
- Phoenix Recursive Feedback — Every completed cycle produces a 3-sentence summary stored for the next higher-resolution compression.
- Native Peer Review Engine — Evidence labeling + mission-spine checks are non-negotiable sub-routines.
- Full Observability — Correlation ID + structured logs + metrics on every multi-agent run.
Activation Protocol (Copy-Paste Ready)
At the start of any complex session, insert:
