The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the open command to launch the generated HTML report in the user's browser. It also anticipates using command-line tools for platforms like Jira, Linear, and GitHub to collect task and issue data.
[DATA_EXFILTRATION]: The skill is designed to search for and read sensitive project data, including architecture descriptions, strategies, operational rules, and configuration templates (e.g., .env.example). While the data remains local, its aggregation into a single HTML report creates a centralized data exposure point.
[PROMPT_INJECTION]: The skill processes numerous untrusted external sources, such as project READMEs, architecture files (CLAUDE.md), and issue descriptions, which can be vectors for indirect prompt injection.
Ingestion points: Step 2 in SKILL.md defines a broad scanning process for project files and CLI outputs from task managers.
Boundary markers: There are no instructions for using delimiters or boundary warnings when interpolating file content into the agent's context.
Capability inventory: The skill allows the agent to perform file system operations (ls, glob), read file contents, and execute shell commands (open).
Sanitization: No data validation or sanitization routines are specified before the information is processed or included in the HTML report.
[EXTERNAL_DOWNLOADS]: The HTML report specification (references/html-design-spec.md) includes references to external fonts hosted on Google Fonts and provides a clickable link to the author's GitHub repository for the skill's source code.

product-data-audit