haxaml 0.5.0

Deterministic agent-management framework that implements FRAME governance for AI-assisted software development.

Haxaml

Haxaml is an LLM-first governance layer for coding agents.

It is an MCP-first project: agents are expected to work through the Haxaml MCP server, with the CLI available mainly for local setup and fallback.

Latest release: 0.5.0.

Haxaml gives agents a deterministic project memory model called FRAME, plus MCP tools that make context, rules, verification, and handoff explicit during real work.

Why It Exists

Agent instructions are usually scattered across prompt files, chat history, local conventions, and whatever the current model remembers. That works until the project grows, the session rolls over, or a different agent enters the repo.

Haxaml keeps the operational truth in versioned project files and exposes it through a predictable workflow. Agents can ask for the right context, follow project rules, verify before claiming success, and record what changed.

What Agents Get

• Project facts, rules, history, expectations, and impact maps in .haxaml/
• Task-specific context packs instead of giant prompt dumps
• Validation and reconcile checks before state is trusted
• Verify/record gates for governed work
• Export paths for native agent files such as AGENTS.md, CLAUDE.md, Cursor rules, Copilot instructions, and Gemini guidance

Install

uvx haxaml-mcp

For persistent local installs:

uv tool install haxaml-mcp

MCP Start

Configure your MCP client to launch haxaml-mcp with HAXAML_PROJECT_DIR set to the project root. See MCP.md for the human/operator guide.

Once connected, agents can initialize and validate through MCP tools:

• haxaml_init
• haxaml_validate

Optional CLI fallback for local setup:

haxaml init
haxaml validate

Bootstrap Prompt

Paste this into your native agent instruction file (AGENTS.md, CLAUDE.md, GEMINI.md, .github/copilot-instructions.md, etc.):

This repository uses Haxaml for agent governance.

Use the Haxaml MCP server for governed project work.
Before governed project work, call haxaml_about(project_dir='.') once in the active MCP session.
Follow the workflow returned by that tool.
Do not edit .haxaml/* for utility or side tasks that are not governed project work.

FRAME Files

• .haxaml/facts.yaml - project truth
• .haxaml/rules.yaml - agent operating rules
• .haxaml/acts.yaml - execution diary and decisions
• .haxaml/expect.yaml - run plan and milestones
• .haxaml/map.yaml - optional module ownership and impact map

Docs

• FRAME.md - FRAME memory model
• haxaml.md - how Haxaml makes FRAME operational
• MCP.md - MCP setup and operator guide
• assurance.md - implementation evidence and current maturity notes
• LONGTERM.md - roadmap direction
• docs/architecture.md - module layout and MCP split overview
• docs/mcp-tool-reference.md - compact MCP tool and resource index
• CONTRIBUTING.md - contributor workflow and expectations
• examples/minimal-governed-flow - minimal FRAME project for governed-flow smoke tests