aigon-research-autopilot

name	aigon-research-autopilot
description	Research autopilot <ID> [agents...] - Fleet setup + spawn + monitor + evaluate

aigon-research-autopilot

Run a fully autonomous Fleet research pipeline. Sets up findings files, spawns agents in parallel, monitors progress, and optionally runs synthesis when all agents submit.

aigon research-autopilot $ARGUMENTS [agents...]

Argument Resolution

If no ID is provided, or the ID doesn't match an existing topic in backlog or in-progress:

1. List all files in ./docs/specs/research-topics/02-backlog/ and 03-in-progress/ matching research-*.md
2. If a partial ID or name was given, filter to matches
3. Present the matching topics and ask the user to choose one

Usage

Basic usage (use configured default agents)

aigon research-autopilot $ARGUMENTS

Specify agents explicitly

aigon research-autopilot $ARGUMENTS cc gg cx

With auto-eval

aigon research-autopilot $ARGUMENTS cc gg cx --auto-eval

What it does

1. Setup phase: If no findings files exist, runs research-start to create them
2. Spawn phase: Creates a tmux session for each agent, running research-do
3. Monitor phase: Polls agent findings files every 30s for complete status, prints status table
4. Evaluate phase: If --auto-eval, runs research-eval automatically; otherwise prints the command

Subcommands

aigon research-autopilot status $ARGUMENTS    # Check agent statuses
aigon research-autopilot stop $ARGUMENTS      # Stop all agents

Options

• --auto-eval — Automatically run research-eval when all agents submit
• --poll-interval=N — Seconds between status checks (default: 30)

Requirements

• tmux must be installed (brew install tmux)
• At least 2 agents must be specified (or configured as defaults)
• Research topic must be in backlog or in-progress

Agent Workflow

Each spawned agent will:

1. Run research-do to conduct their research
2. Write findings to their findings file
3. Run agent-status research-complete to signal completion

Prompt Suggestion

After all agents submit, suggest the eval command:

aigon-research-eval $ARGUMENTS

ARGUMENTS: $ARGUMENTS