The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary function is to guide the agent in executing the asc CLI tool to interact with the Apple Ads API, which involves significant shell command execution capabilities.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and act upon data retrieved from the Apple Ads API, which is an external and potentially untrusted source.
Ingestion points: The agent reads output from asc ads apps search, asc ads campaigns find, and various report commands.
Boundary markers: There are no instructions for the agent to use delimiters or ignore embedded instructions when parsing these API responses.
Capability inventory: The agent has the ability to perform resource mutations such as create, update, delete, pause, and resume based on its interpretation of the ingested data.
Sanitization: No sanitization or validation of the external API content is defined in the skill instructions.
[PRIVILEGE_ESCALATION]: The skill documentation includes the use of the ASC_ADS_BYPASS_KEYCHAIN=1 flag. This environment variable is used to bypass interactive security prompts from the host keychain, which is a common requirement for automated CI/CD pipelines but represents a deliberate reduction in security controls.
[DATA_EXFILTRATION]: The skill provides instructions for accessing and configuring sensitive authentication files, such as the private key at ~/.asc/apple-ads-private-key.pem. While this is standard for the tool's operation, it involves the agent interacting with highly sensitive credential materials.

asc-apple-ads