The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of dynamic shell execution. In its preamble and throughout the instructions, it uses eval and source with process substitution to execute logic generated by local binaries (e.g., eval "$(~/.claude/skills/gstack/bin/gstack-slug)" and source <(~/.claude/skills/gstack/bin/gstack-repo-mode)").
[COMMAND_EXECUTION]: The skill automates project configuration changes by offering to append 'skill routing' rules to the CLAUDE.md file and subsequently performing git operations (git add and git commit) to persist those changes.
[DATA_EXFILTRATION]: The skill implements telemetry and remote synchronization features. It tracks skill usage via a local telemetry binary and offers an 'Artifacts Sync' feature to push project data (plans, designs, and reports) to a private GitHub repository. These mechanisms are prompted to the user but represent paths for data to leave the local environment.
[EXTERNAL_DOWNLOADS]: The skill performs automated network requests for update checks and repository synchronization. It executes gstack-update-check and manages artifact repositories using git fetch and git merge commands.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. Because its primary function is to scrape and ingest arbitrary web content into the agent's context, it creates a vulnerability where malicious instructions embedded in a web page could influence the agent's behavior.

scrape