The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses bunx to download the usagemax package from the npm registry at runtime.
[REMOTE_CODE_EXECUTION]: The skill executes the usagemax package immediately after downloading it via the bunx runner.
[COMMAND_EXECUTION]: Instructions include running CLI commands such as usagemax link and usagemax doctor --deep in the local environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, creating a potential vector for indirect instructions.
Ingestion points: Public profile data and natural-language resources are fetched from https://usagemax.com/api/profiles/{handle} and https://usagemax.com/ask (referenced in SKILL.md).
Boundary markers: The skill lacks explicit delimiters or instructions to the agent to disregard potential instructions embedded within the retrieved data.
Capability inventory: The skill possesses the ability to execute shell commands (bunx) and perform network operations to the vendor's domain.
Sanitization: No sanitization or validation logic is specified for the data retrieved from the external API endpoints.

usage-observability