Skip to main content
The Rosetta is the core of AXL Protocol. 445 lines. Any LLM reads it once, becomes fluent. No training. No fine-tuning.

​
What It Contains

SectionsContent
1-6Packet shape, operations, tags, symbols, relations, tokenizer profile
7Four bridges (temporal, spatial, logical, epistemic)
8Operation schemas with field patterns
9Cross-domain examples (finance, medicine, military)
107-step generation rules
11Backward compatibility (v1.0, v2.1, v2.2)
12-13Payment proof, encoding layers
14The P-S-A-S-E-G-D algorithm
15Directive (activates behavioral change)
16v2.2 preamble manifests (PROFILE, AUDIENCE, ONTOLOGY, BUNDLE)
17Canonical decompression
18Full v2.2 exchange example

​
The Teaching Algorithm

The Rosetta follows P-S-A-S-E-G-D:
  1. Prime — activates instruction-following mode
  2. Shape — shows the template before details
  3. Alphabet — key-value associations in tables
  4. Schemas — type information in positional templates
  5. Examples — minimum viable training set (4 domains x 7 ops)
  6. Generate — teaches output production, not just parsing
  7. Direct — activates behavioral change

​
Fetch It

The live Rosetta: axlprotocol.org/rosetta 
curl -s https://axlprotocol.org/rosetta

import urllib.request
rosetta = urllib.request.urlopen("https://axlprotocol.org/rosetta").read().decode()