The Agent Skills Directory

[Prompt Injection] (SAFE): No malicious override or bypass markers were detected. The skill's instructions specifically include a safeguard in the 'Missing-info strategy' and 'Checklists' to not request or record secrets or credentials.
[Data Exposure & Exfiltration] (SAFE): There are no commands that access sensitive files (e.g., ~/.ssh, .env) or perform network operations to external domains.
[Remote Code Execution] (SAFE): The skill does not contain any patterns for downloading or executing remote scripts or external packages.
[Indirect Prompt Injection] (LOW): The skill processes user-supplied context to populate its planning templates. * Ingestion points: User context about products and teams provided via the primary prompt and intake questions in references/INTAKE.md. * Boundary markers: Absent; the skill relies on the structure of the templates to isolate data. * Capability inventory: Limited to writing markdown files to a user-specified directory as described in README.md. * Sanitization: None performed on the input text before inclusion in templates.
[Obfuscation] (SAFE): No Base64, zero-width characters, homoglyphs, or other encoding techniques used to hide content were found.
[Privilege Escalation] (SAFE): No commands for elevating privileges or modifying system configurations are present.

design-engineering