gman - Universal Credential Manager

gman is a command-line tool designed to streamline credential and secret management for your scripts, automations, and applications. It provides a single, secure interface to store, retrieve, and inject secrets, eliminating the need to juggle different methods like configuration files or environment variables for each tool.

Overview

The core philosophy of gman is to act as a universal wrapper for any command that requires credentials. You can store your secrets—like API tokens, passwords, or certificates—in an encrypted vault backed by various providers. Then, you can either fetch them directly or, more powerfully, execute commands through gman, which securely injects the necessary secrets as environment variables or command-line flags.

Features

• Secure, Encrypted Storage: All secrets are stored in an encrypted state using strong cryptography.
• Pluggable Providers: Supports different backends for secret storage. The default is a local file-based system.
• Git Synchronization: The local provider can synchronize your encrypted secrets across multiple systems using a private Git repository.
• Seamless Command Wrapping: Run any command through gman to automatically provide it with the secrets it needs (e.g., gman aws s3 ls).
• Customizable Run Profiles: Define how secrets are passed to commands, either as environment variables (default) or as specific command-line flags.
• Secret Name Standardization: Enforces snake_case for all secret names to ensure consistency.
• Direct Secret Access: Retrieve plaintext secrets directly when needed (e.g., gman get my_api_key).
• Dry Run Mode: Preview the command and the secrets that will be injected without actually executing it using the --dry-run flag.

Installation

Cargo

If you have Cargo installed, then you can install gman from Crates.io:

cargo install gman

# If you encounter issues installing, try installing with '--locked'
cargo install --locked gman

Configuration

gman is configured via a YAML file located somewhere different for each OS:

Linux

$HOME/.config/gman/config.yml

Mac

$HOME/Library/Application Support/gman/config.yml

Windows

%APPDATA%/Roaming/gman/config.yml

Default Configuration

---
provider: local
password_file: ~/.gman_password

# Optional Git sync settings for the 'local' provider
git_branch: null # Defaults to 'main'
git_remote_url: null # Required for Git sync
git_user_name: null # Defaults to global git config user.name
git_user_email: null # Defaults to global git config user.email
git_executable: null # Defaults to 'git' in PATH
run_configs: null # List of run configurations (profiles)

Provider: local

The default local provider stores an encrypted vault file on your filesystem. For use across multiple systems, it can sync with a remote Git repository.

Important Notes for Git Sync:

• You must create the remote repository on your Git provider (e.g., GitHub) before attempting to sync.
• The git_remote_url must be in SSH or HTTPS format (e.g., git@github.com:your-user/your-repo.git).

Example local provider config for Git sync:

provider: local
git_branch: main
git_remote_url: "git@github.com:my-user/gman-secrets.git"
git_user_name: "Your Name"
git_user_email: "your.email@example.com"

Run Configurations

Run configurations (or "profiles") tell gman how to inject secrets into a command. When you run gman <command>, it looks for a profile with a name matching <command>. If found, it injects the specified secrets. If no profile is found, gman will error out and report that it could not find the run config with that name.

Important: Secret names are always injected in UPPER_SNAKE_CASE format.

Basic Run Config (Environment Variables)

By default, secrets are injected as environment variables. The two required fields are name and secrets.

Example: A profile for the aws CLI.

run_configs:
  - name: aws
    secrets:
      - aws_access_key_id
      - aws_secret_access_key

When you run gman aws ..., gman will fetch these two secrets and expose them as environment variables to the aws process.

Advanced Run Config (Command-Line Flags)

For applications that don't read environment variables, you can configure gman to pass secrets as command-line flags. This requires three additional fields: flag, flag_position, and arg_format.

• flag: The flag to use (e.g., -e).
• flag_position: An integer indicating where to insert the flag in the command's arguments. 1 is immediately after the command name.
• arg_format: A string that defines how the secret is formatted. It must contain the placeholders {key} and {value}.

Example: A profile for docker run that uses the -e flag.

run_configs:
  - name: docker
    secrets:
      - my_app_api_key
      - my_app_db_password
    flag: -e
    flag_position: 2 # In 'docker run ...', the flag comes after 'run', so position 2.
    arg_format: "{key}={value}"

When you run gman docker run my-image, gman will execute a command similar to: docker run -e MY_APP_API_KEY=... -e MY_APP_DB_PASSWORD=... my-image

Usage

Storing and Managing Secrets

All secret names are automatically converted to snake_case.

• Add a secret:

  # The value is read from standard input
  echo "your-secret-value" | gman add my_api_key
  

  or don't provide a value to add the secret interactively:

  gman add my_api_key
  

• Retrieve a secret:

  gman get my_api_key
  

• Update a secret:

  echo "new-secret-value" | gman update my_api_key
  

  or don't provide a value to update the secret interactively:

  gman add my_api_key
  

• List all secret names:

  gman list
  

• Delete a secret:

  gman delete my_api_key
  

• Synchronize with remote secret storage (specific to the configured provider):

  gman sync
  

Running Commands

• Using a default profile:

  # If an 'aws' profile exists, secrets are injected.
  gman aws sts get-caller-identity
  

• Specifying a profile:

  # Manually specify which profile to use with --profile
  gman --profile my-docker-profile docker run my-app
  

• Dry Run:

  # See what command would be executed without running it.
  gman --dry-run aws s3 ls
  # Output will show: aws -e AWS_ACCESS_KEY_ID=***** ... s3 ls
  

Creator

• Alex Clarke