The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by processing raw user input without sanitization or boundary markers.
Ingestion points: The $ARGUMENTS variable in SKILL.md receives raw, unvalidated input from the user.
Boundary markers: Absent. The input is interpolated directly into instructions for the agent to split and use in shell commands without delimiters or instructions to ignore embedded commands.
Capability inventory: The skill uses the Bash tool to execute uvx cookiecutter and uv sync (SKILL.md).
Sanitization: Absent. There is no logic to filter or escape the content of $ARGUMENTS before it influences the command construction.
[COMMAND_EXECUTION]: The skill uses the Bash tool to perform system-level operations for project scaffolding.
Evidence: uvx cookiecutter --no-input ${CLAUDE_SKILL_DIR}/assets/project-template ... and uv sync in SKILL.md.
Context: These commands are used to automate the setup of a Scrapy development environment, which is the primary purpose of the skill.
[EXTERNAL_DOWNLOADS]: The skill triggers the download of Python packages and CLI utilities from official registries.
Evidence: uv sync installs several Scrapy-related dependencies, and uvx cookiecutter downloads the cookiecutter tool.
Context: The dependencies (e.g., scrapy, scrapy-zyte-api) are well-known libraries in the scraping ecosystem. Several are vendor-maintained resources from 'zytedata', consistent with the skill's authorship.

scrape-ensure-project