The Agent Skills Directory

[DATA_EXFILTRATION]: The skill uses curl to make network requests to external endpoints, such as https://staging.example.com/api/chat, for testing rate-limiting behavior. These requests target domains outside the whitelisted set.
[COMMAND_EXECUTION]: The skill executes shell commands, including recursive grep to scan the application directory for specific keywords and curl within loops to simulate high-traffic burst requests.
[PROMPT_INJECTION]: The skill scans local project files to build an inventory of endpoints, which constitutes an indirect prompt injection surface where external data could influence the agent's context.
Ingestion points: Scanning the app/api directory using grep for route identification.
Boundary markers: None present; the results of file scans are processed without delimiters or instructions to ignore embedded content.
Capability inventory: The skill has file-read capabilities (grep) and network-write capabilities (curl) across the audit workflow.
Sanitization: No sanitization or validation is applied to the data extracted from the filesystem before it is used in the audit report.

rate-limit-audit