my_skill

Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in (single or multi-account), or reading/injecting/...

This skill appears to do what it says (install and run 1Password CLI), but there are two things to check before installing: (1) The runtime instructions require tmux but the skill did not declare tmux as a required binary. If you plan to use it, ensure tmux is available and the skill metadata is updated to list tmux as a dependency. (2) The sample flow captures tmux pane output (tmux capture-pane) which can accidentally include secrets printed by commands; the skill does not show filtering or sanitization. Ask the publisher to (a) declare tmux as a required binary, (b) avoid capturing raw pane output that may contain secrets or explicitly filter/redact it, and (c) document how captured output is used and where it may be stored or displayed. Also verify you trust the Homebrew 1password-cli formula and only run these automated flows interactively (do not allow unattended/automated sign-ins without reviewing outputs). If you are uncomfortable with those issues, run the official 1Password CLI manually instead of using this skill.

Type: OpenClaw Skill Name: myskilltest Version: 1.0.0 The skill is designed to integrate with the 1Password CLI, which is a legitimate purpose. However, it contains strong, mandatory instructions (prompt injection) for the AI agent to *always* execute `op` commands within a dedicated `tmux` session, as detailed in `SKILL.md`. While the stated reason for this is to 'avoid re-prompts and failures,' this level of control over the agent's execution environment for a sensitive tool like 1Password CLI represents a risky capability. There is no clear evidence of intentional malicious behavior like data exfiltration or backdoor installation, and the skill includes guardrails for secure secret handling. The `tmux` requirement, though, is a significant and unusual directive that could be a vector for future abuse or unintended behavior if the agent's `tmux` handling or the `tmux` skill itself were compromised.