Skip to content
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
76 changes: 76 additions & 0 deletions docs/quickstart.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,76 @@
```markdown
# Consensus Hardening Protocol (CHP) – Quickstart

Consensus Hardening Protocol (CHP) is an open-source multi-agent decision orchestration tool (Python 3.10+, MIT licensed) that generates structured, verifiable consensus reports with executable workflows.

## Prerequisites
- Python 3.10 or later
- Git (to clone the repository)

## Installation
```bash
git clone https://github.com/Cubiczan/consensus-hardening-protocol.git
cd consensus-hardening-protocol
pip install -e .
```

Once installed, the command-line tool `cme` becomes available. Verify it by running a plain demo:

```bash
cme demo
```

## Run without installing
If you prefer not to install the package, you can run the same demo directly:

```bash
PYTHONPATH=src python3 -m cme.cli demo
```

## Run a basic consensus round
A basic consensus round is simply a demo run that includes a decision question. After installation, execute:

```bash
cme demo "Should we invest $4M in a new enterprise tier next quarter?"
```

This single command triggers a full multi-agent deliberation pipeline and writes an orchestration report.

## What to expect
The command above generates a complete Markdown orchestration report. The report includes the following sections:

- **Problem classification**
- **Per-agent reasoning traces**
- **Grounding verdicts**
- **Playbook deltas**
- **Final executable workflow**

The demo runs three agents in sequence:
1. **Finance agent** runs first and produces its conclusions.
2. **Strategy agent** reads the finance agent’s output and builds on it.
3. **Compliance agent** reads both previous outputs and issues a conditional approval.

Finally, a **synthesizer** takes all agent contributions and produces a **Statement** and an **executable Workflow**.

The output is written as a Markdown file – look for the generated report in your working directory.

## Further reading
- Core state models: `src/cme/chp/`
- Complete demonstration script: `DEMO_SCRIPT.md`
- Finance-focused roadmap: `docs/CHP_FINANCE_PROJECT_ROADMAP.md`
- Minimal example file: `examples/basic_demo.py`
- Run the test suite (42 tests passing):
```bash
pip install pytest
PYTHONPATH=src pytest tests/ -v
```
- Explore initial agent playbooks and context:
```bash
cme playbook finance
cme context
```
- Launch a CHP funding allocation session:
```bash
cme chp-start
```
```