The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the agent to execute a shell command via the bash tool as a mandatory first step for every user request.- [EXTERNAL_DOWNLOADS]: The skill performs a runtime network operation using curl to fetch content from an external URL (https://example.com/junior-eval-oauth/whoami).- [INDIRECT_PROMPT_INJECTION]: The skill ingests external content from a network tool into the agent's context, creating a potential vulnerability surface for indirect instructions.
Ingestion points: The output of the curl command executed in the bash tool (referenced in SKILL.md).
Boundary markers: Absent; there are no instructions to wrap or delimit the external tool output.
Capability inventory: The skill uses the bash tool, which provides high-level system interaction capabilities.
Sanitization: Absent; the skill does not specify any validation or filtering of the content received from the external URL.

eval-oauth