The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (SAFE): The skill references standard installation methods (npm, Homebrew, Go) for the 'beads' tool. It explicitly instructs the agent to ask for user confirmation before proceeding with global system package installations, which is a security best practice.
[COMMAND_EXECUTION] (SAFE): All command-line examples (bd, git, npm, brew, pip) are directly related to the stated purpose of the skill (task management and version control). No suspicious or obfuscated commands were detected.
[REMOTE_CODE_EXECUTION] (SAFE): While the skill mentions installing a tool from GitHub (github.com/steveyegge/beads), the repository belongs to a known developer and the installation uses the standard go install mechanism. No piped-to-shell or unverified remote script executions are present.
[PROMPT_INJECTION] (SAFE): The 'CRITICAL' and 'IMPORTANT' markers used in the documentation are strictly for operational guidance (e.g., ensuring git sync before ending a session) and do not attempt to bypass AI safety filters or override system instructions.
[DATA_EXFILTRATION] (SAFE): The skill interacts with the local .beads directory and the project's git repository. There are no patterns suggesting the exfiltration of sensitive files or credentials to external domains.

beads