The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides commands for the user to install the Claude Code tool using remote scripts hosted at https://claude.ai/install.sh and https://claude.ai/install.ps1. As these originate from the official domain of the service provider, they are documented as trusted installation vectors.
[EXTERNAL_DOWNLOADS]: The tutorial guides users to install various plugins and add marketplaces using commands like /plugin install. It references external repositories such as obra/superpowers-marketplace and team-attention/plugins-for-claude-natives. The latter is explicitly identified as a resource created by the skill author's organization for instructional purposes.
[COMMAND_EXECUTION]: The skill includes instructions to modify the user's environment, such as changing the Windows PowerShell execution policy (Set-ExecutionPolicy) and updating local configuration files (~/.claude/settings.json and ~/.claude/settings.local.json). These actions are transparently presented to the user as part of the tool's setup process.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch official documentation using curl and read it to provide accurate answers if its internal knowledge is insufficient. This creates a surface for processing external data.
Ingestion points: Network fetch of documentation URLs into local files for subsequent reading (SKILL.md).
Boundary markers: None explicitly implemented for the external content.
Capability inventory: Includes subprocess execution for network requests and file system access across the referenced files.
Sanitization: No specific sanitization or filtering logic for the fetched documentation is described.

day1-onboarding