The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill operates on existing project files and takes user input to determine project structure, creating an attack surface for indirect prompt injection if the processed data contains malicious instructions.
Ingestion points: The skill uses Read and Glob tools to analyze existing project structures and AskUserQuestion to gather requirements from the user (SKILL.md).
Boundary markers: No explicit boundary markers or instructions to ignore embedded content are defined for the data ingestion phase.
Capability inventory: The agent has access to Bash for command execution and Write/Edit for file system modification (SKILL.md).
Sanitization: There is no explicit mention of sanitization or validation of the data read from the local environment before it is used to influence the agent's planning or execution.
[COMMAND_EXECUTION]: The skill explicitly uses the Bash tool to perform project initialization tasks such as git init and npm init (SKILL.md).
Evidence: The "Tools Used" section lists Bash for initializing tools. This is a standard and expected capability for a project bootstrapping utility.

project-bootstrapper