Agent Machine Language · Static validation

AML Studio

AML Studio is a developer interface for submitting Agent Machine Language to the HELLO OS static parser. It displays parse results, an abstract syntax tree, intent graph nodes, declared budgets, token count, diagnostics, and SHA-256 content hashes.

Run the interactive studio

Start the HELLO OS application locally with Docker Compose, then open http://localhost:3000/aml-studio. The interface sends the source to POST /api/aml/compile. The endpoint accepts a JSON object with a string source field and limits input to 64 KB.

git clone https://github.com/CodesbyFebin/Hello-OS.git
cd Hello-OS
docker compose up --build

Call the validator directly

curl -sS http://localhost:3000/api/aml/compile \
  -H 'content-type: application/json' \
  -d '{"source":"mission example {\n  steps {\n    verify: proof.check($input) → $result\n  }\n}"}'

A successful response includes ok, executable: false, the parsed AST and intent graph, declared capabilities and budgets, diagnostics, source and graph SHA-256 values, and a lexical token count. Invalid input returns diagnostics with HTTP 422. The token count is a parser count, not a model billing estimate.

Example and implementation references

• AML examples and API format
• Current AML grammar
• Static parser implementation
• Compile endpoint implementation
• Parser tests

Mission Control and real data

The uploaded Mission Control HTML is a visual prototype containing randomized example metrics and generated log entries. HELLO OS does not present those values as live system data. The application’s Ecosystem dashboard instead requests current API-backed counts and activity from the configured database. In a fresh installation, values may be zero or empty until the workspace has records.

After starting the app, open http://localhost:3000/ecosystem to inspect that API-backed view. A live view depends on the app server and database being available.