Skip to main content

Documentation Index

Fetch the complete documentation index at: https://bastani.mintlify.app/llms.txt

Use this file to discover all available pages before exploring further.

Atomic ships three built-in workflows. Run them with atomic workflow -n <name> -a <agent> "<prompt>". They are reserved — a local or global workflow with the same name will not shadow a built-in when running via the atomic CLI.

​
ralph

The Ralph Method enables multi-hour autonomous coding sessions. Approve a spec, let Ralph work in the background, and check back later. How it works. A planner sub-agent decomposes the spec into a task list with dependency tracking, stored in SQLite (WAL mode for parallel access). An orchestrator retrieves the list, validates the dependency graph, and dispatches worker sub-agents for ready tasks. A reviewer audits implementations with structured JSON output; on P0–P2 findings, a debugger investigates root causes and feeds back to the planner on the next iteration. Loop config. Up to 10 iterations. Exits early after 2 consecutive clean reviews (zero actionable findings). P3 (minor) findings are non-actionable. 
# From a prompt
atomic workflow -n ralph -a claude "Build a REST API for user management"

# From a spec file
atomic workflow -n ralph -a claude "specs/2026-05-12-my-feature.md"

# Detached so you can do other work — attach later
atomic workflow -n ralph -a claude -d "build the auth module"
atomic workflow session connect atomic-wf-claude-ralph-<id>
Run Ralph in a git worktree so autonomous changes stay isolated from your working tree.
git worktree add ../my-project-ralph feature-branch
cd ../my-project-ralph
atomic workflow -n ralph -a claude "Build the auth module"
Ralph runs on all three agents: -a claude, -a opencode, or -a copilot.

​
deep-research-codebase

A multi-agent parallel research pipeline. Not a one-shot command — a full pipeline that produces a permanent research artifact.
  1. Scout — one agent scans codebase structure and writes an architectural orientation.
  2. History — a parallel agent surfaces prior research from research/docs/.
  3. Explorers — multiple parallel agents (count scaled by LOC) each investigate a partition.
  4. Aggregator — a final agent synthesizes explorer reports + history into a dated doc at research/docs/YYYY-MM-DD-<slug>.md.
atomic workflow -n deep-research-codebase -a claude "How does the authentication system work?"
The output persists in research/docs/, so future runs, specs, and workflows can reference it. Pair it with the /create-spec skill (see context files) to turn a research doc into an execution spec for Ralph.

​
open-claude-design

Launches an Open Claude Design session — produces a designed prototype or artifact from a structured prompt. Uses declared inputs, so flags map one-to-one onto schema fields: 
atomic workflow -n open-claude-design -a claude \
  --prompt="a dashboard for monitoring API latency" \
  --output-type=prototype
Run atomic workflow inputs open-claude-design -a claude to see the full schema, or omit -n to fill the form in the interactive picker.

​
Listing built-ins

atomic workflow list                 # everything, grouped by source
atomic workflow list -a claude       # filter by agent
Built-ins always appear first, followed by any custom workflows you’ve registered.

Running workflows

Picker, structured inputs, detach, and reading run artifacts.

Building your own

Author a workflow with the SDK.