The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because its core function involves reading and analyzing untrusted source code files from the user's project (specifically .swift files). An attacker could embed malicious instructions in code comments or strings intended to mislead the agent during the audit process.
Ingestion points: Project source code files read via the Read, Grep, and Glob tools across all audit layers.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions embedded within the files it processes.
Capability inventory: The agent has access to powerful tools like Bash, Edit, and Write, which increases the potential impact if an indirect injection successfully influences the agent's behavior.
Sanitization: The skill does not implement sanitization or validation of file content before it is processed by the AI.
[EXTERNAL_DOWNLOADS]: The skill's shared configuration (radar-suite-core.md) recommends that users install a third-party tool named 'Dippy' from a personal GitHub repository (ldayton/dippy) if their project path contains spaces. While this is a recommendation to the user and not an automated execution by the skill, it points to unverified external code from an untrusted source.

ui-path-radar