The Agent Skills Directory

[DATA_EXFILTRATION]: The skill reads the contents of local skill directories and uploads them to the Volcengine security scanning service at open.volcengineapi.com. This is the primary documented purpose of the skill, and the network destination is a vendor-owned resource associated with the skill's author (bytedance).
[COMMAND_EXECUTION]: The skill executes a Python script (scripts/scan.py) that performs local file system operations, including recursively zipping directories, and makes authenticated network requests to the Volcengine API using provided environment variables.
[INDIRECT_PROMPT_INJECTION]: The skill acts as an ingestion surface by reading and processing content from other skills in the workspace for auditing purposes.
Ingestion points: The scripts/scan.py script reads all files within a user-specified directory via the --path argument.
Boundary markers: None implemented for the ingestion phase.
Capability inventory: The skill has the capability to read local files, create temporary zip archives, and transmit data to external ByteDance/Volcengine infrastructure.
Sanitization: The skill packages the ingested data as a binary archive for remote analysis rather than interpreting or executing the instruction content locally.

byted-security-skillsscanner