The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute shell commands using the Bitwarden Secrets Manager CLI (bws).
Evidence: The skill defines operations for bws secret list, bws secret get, and bws secret create.
Evidence: The skill includes the bws run -- '<command>' functionality, which executes a process with secrets injected into its environment.
Security Controls: The skill explicitly warns the user and the agent to only run trusted commands, noting that bws run executes in the shell and could allow malicious scripts to access injected secrets.
[SAFE]: The skill adheres to safe handling procedures for sensitive operations.
Evidence: It requires the agent to obtain explicit user confirmation before executing any "Write Operations," such as creating, editing, or deleting secrets and projects.
Evidence: It instructs the agent to proactively check the Bitwarden vault for existing credentials before asking the user to provide them, reducing the frequency of sensitive data handling in the chat context.
[DATA_EXFILTRATION]: No suspicious data transfer patterns were identified.
Analysis: Although the skill manages credentials (API keys, tokens, passwords), it does not include instructions to send this data to external or untrusted third-party domains. All network activity is confined to the standard behavior of the Bitwarden CLI tool.

bitwarden-secrets