crewkit
MarketplaceFeaturesPricingDocsGitHub
Sign InGet Started
crewkitInstallationQuickstartConfigurationTroubleshooting

Usage

TUICLIDashboardAnalytics

Configure

AgentsSkillsRulesCommandsInheritancePlaybooksExperimentsHooksTeamWorkspacesFAQ

API Reference

API OverviewAuthenticationSessions APIResources APIOrganizations APIProjects APIPlaybooks APIExperiments API

Configuration

Configure crewkit with environment variables, config files, and CI/CD settings.

crewkit is configured through environment variables and local config files. Most settings have sensible defaults — you only need to change things for CI/CD or multi-environment setups.


Environment variables

VariableDescriptionDefault
CREWKIT_API_URLAPI base URLhttps://api.crewkit.io
CREWKIT_ORGOrganization slug (skips auto-detection)—
CREWKIT_REPOSITORY_IDRepository ID (skips detection)—
CREWKIT_ENVEnvironment name (used for error reporting)production
NO_COLORDisable colored output—

Local config file

When you run crewkit init or crewkit code for the first time in a repo, crewkit creates a .crewkit/config.toml file (gitignored):

version = 1

[repository]
id = "repo_abc123"
remote_url = "git@github.com:acme/frontend"
linked_at = "2026-01-20T10:30:00Z"

[component]
detected = "dashboard"
override = null

[sync]
last_sync = "2026-01-20T10:35:00Z"
agents = ["frontend-expert", "api-designer"]

This file caches your repo-to-project mapping so subsequent runs skip the detection step.

If you move your repo or change the git remote, crewkit invalidates the cache and re-prompts.


CI/CD mode

For non-interactive environments (GitHub Actions, GitLab CI, etc.), combine --yes with environment variables:

# Set in your CI environment
export CREWKIT_REPOSITORY_ID=repo_abc123

# Run without prompts
crewkit code --yes --prompt "Run the test suite and fix failures"

crewkit auto-detects non-TTY environments and provides a helpful error if interactive prompts are needed but unavailable.


Headless mode

Run crewkit without the TUI for scripting and automation:

# Run a one-shot prompt
crewkit code --prompt "Add error handling to api/auth.rs"

# JSON output for parsing
crewkit code --prompt "List all TODO comments" --output-format json

# Stream JSON for real-time processing
crewkit code --prompt "Refactor the user model" --output-format stream-json

# Set a timeout (default: 300 seconds)
crewkit code --prompt "Fix the failing test" --timeout 600

Display options

# Disable colors
crewkit --no-color code

# ASCII-only output (no unicode symbols)
crewkit --plain code

# Verbose output (file lists, timing details)
crewkit code --verbose

Debug logging

Enable debug logs for troubleshooting:

crewkit code --debug

Logs are written to .crewkit/debug-latest.log in your project root. Follow them in another terminal:

tail -f .crewkit/debug-latest.log

Next steps

  • Troubleshooting
  • CLI command reference

Quickstart

Go from install to your first tracked session in under a minute.

Troubleshooting

Common issues and how to fix them.

On this page

Environment variablesLocal config fileCI/CD modeHeadless modeDisplay optionsDebug loggingNext steps
crewkit

Observability, governance, and continuous improvement for AI-assisted engineering teams.

Product

  • Marketplace
  • Features
  • Pricing
  • Documentation

Resources

  • GitHub
  • Changelog
  • Report Issue
  • System Status

Company

  • About
  • Privacy
  • Terms

© 2026 Karibew Inc. All rights reserved.

Command Palette

Search conversations, projects, playbooks, and more