The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses WebFetch and the context7 tool to retrieve documentation and SDK information from external URLs provided by the user or identified via the tool name.
[COMMAND_EXECUTION]: The skill generates and executes shell commands for various purposes, including configuration of environment variables via the Vercel (vercel env add) and Heroku (heroku config:set) CLIs, as well as performing smoke tests using curl and node -e to verify the functionality of the integration.
[DATA_EXFILTRATION]: The skill handles sensitive credentials such as API keys and secrets; however, it includes strong security mitigations. It explicitly instructs the agent to use the --sensitive flag for secret variables in Vercel to prevent them from being readable in local environments and provides clear warnings against committing sensitive files to version control.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes external documentation. It mitigates this risk by including a mandatory human-in-the-loop summary step where the user must review the proposed changes and configuration before any files are scaffolded or commands are executed.

toolify