The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: Fetches and executes the official installation script for Claude Code from the claude.ai domain.
Evidence: curl -fsSL https://claude.ai/install.sh | sh in SKILL.md.
[COMMAND_EXECUTION]: Utilizes various gitpod CLI commands to manage workspace lifecycle, including project creation, secret configuration, and environment testing.
Evidence: Commands such as gitpod project create, gitpod project secret set, and gitpod environment ssh are described in Phase 2 and Phase 3 of SKILL.md.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data by scanning repository files to detect the tech stack.
Ingestion points: Reads local files such as package.json, Gemfile, and composer.json in Phase 1.1 of SKILL.md to identify frameworks and databases.
Boundary markers: No specific delimiters are defined for reading these repository files.
Capability inventory: The skill has access to shell execution (Bash), file writing (Write), and network-capable CLI tools (gitpod).
Sanitization: No explicit sanitization or filtering of the file contents is mentioned before framework detection logic is applied.

setup-devcontainer