AINL (AI-Native Language) for Hermes

What it does

Provides instructions and tooling to integrate AINL — an AI-native language for deterministic graphs — with the Hermes Agent. It helps install runtime components, import and compile .ainl workflow files, run validated graphs, and expose an ainl-mcp bridge for MCP servers so agents can execute repeatable, checkable workflows.

When to use it

Use when the user needs portable compiled workflows rather than ad-hoc prompt loops: importing workflow definitions, validating deterministic graphs, wiring MCP servers, or creating a closed learning loop where behavior evolution is serialized back into .ainl. Trigger on mentions of Hermes, MCP, compile, import, ainl, or deterministic workflow.

What's included

• Scripts: install.sh and helper bridge scripts (ainl_hermes_bridge.py) in the repo; example .ainl files and example bridges.
• References: README and packaging notes; install and pip-based alternatives documented.
• Instructions: install paths, pip install alternatives, command examples (ainl import/compile/run), and MCP registration guidance.

Compatible agents

Best with agents that can run shell installs and manage MCP servers (Hermes-compatible agents, or any agent with shell/exec capabilities and MCP support).