The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local shell commands including rg (ripgrep) for scanning project files and firebase emulators:exec for running security rule test suites. These are standard developer operations consistent with the skill's stated purpose of auditing Firebase environments.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes external content including security rules and application source code, which constitutes an ingestion surface for potential adversarial instructions.
Ingestion points: firestore.rules, storage.rules, firebase.json, and application source files in src, app, and functions directories.
Boundary markers: The instructions do not define specific delimiters for separating untrusted data from the audit prompt.
Capability inventory: The skill has the ability to execute shell commands (rg, firebase) and perform file system reads.
Sanitization: No explicit sanitization or filtering of the audited content is described.

firebase-security-rules-auditor