The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill automates the installation of various developer utilities by downloading scripts and binaries from established service providers and official repositories. Evidence includes:
- Fetches the official Homebrew installation script from GitHub (package-managers/scripts/install.sh).
- Downloads the Task runner installer from the official domain (make/scripts/install.sh).
- Retrieves the Docker Engine convenience script from Docker's official domain (docker/scripts/install.sh).
- Downloads PowerShell Core installation resources from Microsoft's official domains (powershell-core/scripts/install.sh).
- Fetches tools such as ripgrep and jq directly from their official GitHub release pages (regex/scripts/install.sh, jq/scripts/install.sh).
[COMMAND_EXECUTION]: The installation scripts execute shell commands to manage system packages and configure environment settings. This is standard behavior for developer environment provisioning scripts.
[PRIVILEGE_ESCALATION]: Several scripts utilize sudo to perform administrative tasks, such as installing system-level packages (via apt, dnf, or brew), creating directories in /usr/local/bin, and modifying /etc/shells or group memberships (docker group). These operations are necessary for the stated purpose of setting up system tools.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it facilitates the ingestion of external data from APIs and files using tools like curl, jq, and bash.
- Ingestion points: Data enters the agent's context through curl API responses and file reads (Get-Content, cat) described in curl/SKILL.md, jq/SKILL.md, and bash/SKILL.md.
- Boundary markers: The instructions do not define specific structural delimiters or warnings for the agent when processing potentially untrusted runtime data.
- Capability inventory: The skill provides high capabilities, including full shell scripting access, network operations, and file system modification across multiple scripts.
- Sanitization: The skill emphasizes best practices like variable quoting and using ShellCheck to mitigate common scripting errors, though it does not implement specific sanitization for natural language instructions embedded in processed data.

tools