env - SKILL.md Agent Skill

name: env description: Manage environment variables and secrets with flow (always use Flow env store) source: flow-default

Environment Variables & Secrets

Flow provides a secure way to manage environment variables across projects.

Rules (Must)

Always use Flow env store for secrets/tokens in Flow projects.
Do not rely on shell env vars or .env files for secrets unless they are injected via f env.
For tasks/scripts that need secrets, fetch them with f env get or run via f env run.

Setup

1. Define variables in flow.toml

Add a [storage] section to your project's flow.toml:

[storage]
provider = "myflow.sh"

[[storage.envs]]
name = "local"
description = "Local development"
variables = [
  { key = "DATABASE_URL" },
  { key = "API_KEY" },
  { key = "SECRET_TOKEN", default = "" },
]

2. Set environment variables

Use f env set to store values:

# Set individual env vars
f env set API_KEY=abc123
f env set DATABASE_URL="postgres://..."

# Values are stored in ~/.config/flow/env-local/personal/production.env

3. Pull env vars to local .env

# Pull all env vars for the current environment
f env pull

# Show current env vars
f env list

# Get specific var
f env get API_KEY

Commands

Command	Description
`f env set KEY=value`	Store an env var
`f env pull`	Pull env vars to local .env file
`f env push`	Push local .env to cloud
`f env list`	List env vars for this project
`f env get KEY`	Get specific env var(s)
`f env keys`	Show configured env keys from flow.toml
`f env setup`	Interactive env setup
`f env guide`	Guided prompt to set required vars
`f env run <cmd>`	Run command with env vars injected

Environment Selection

Flow supports multiple environments:

[[storage.envs]]
name = "local"
variables = [{ key = "DATABASE_URL" }]

[[storage.envs]]
name = "staging"
variables = [{ key = "DATABASE_URL" }]

[[storage.envs]]
name = "production"
variables = [{ key = "DATABASE_URL" }]

Example: Spotify CLI

[storage]
provider = "myflow.sh"

[[storage.envs]]
name = "local"
description = "Spotify API credentials"
variables = [
  { key = "SPOTIFY_CLIENT_ID" },
  { key = "SPOTIFY_CLIENT_SECRET" },
  { key = "SPOTIFY_ACCESS_TOKEN" },
  { key = "SPOTIFY_REFRESH_TOKEN", default = "" },
]

Then:

# Set your credentials (example values)
f env set SPOTIFY_CLIENT_ID=example_client_id
f env set SPOTIFY_CLIENT_SECRET=example_client_secret

# Run CLI with env vars injected
f env run bun run src/main.ts now

# Or pull to .env first
f env pull
source .env
bun run src/main.ts now

Task pattern (required for secrets)

When writing Flow tasks, prefer:

MY_TOKEN="$(FLOW_ENV_BACKEND=local f env get --personal MY_TOKEN -f value 2>/dev/null || true)"
if [ -z "${MY_TOKEN:-}" ]; then
  echo "MY_TOKEN missing. Save it with envnew MY_TOKEN=..."
  exit 1
fi
export MY_TOKEN

One-time passwords (1Password Connect)

Use Flow's OTP command to fetch TOTP codes from 1Password Connect:

f otp get <vault> <item> [--field <label>]

Requires:

OP_CONNECT_HOST
OP_CONNECT_TOKEN (env or Flow personal env store)

Storage Locations

Personal env vars: ~/.config/flow/env-local/personal/production.env
Project env vars: Stored via myflow.sh cloud backend
Local .env: Created in project root via f env pull

Authentication

Flow uses a token stored in ~/.config/flow/auth.toml to authenticate. If you haven't authenticated:

f auth login

Security Notes

Personal env vars stored locally in ~/.config/flow/
Project env vars can be synced to cloud via f env push
Never commit .env files to git (add to .gitignore)
Use f env run to inject vars without creating .env files