The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a routing logic that directs the agent to process untrusted external data through various specialized sub-agents. \n- Ingestion points: The routing logic triggers the use of agents like test-runner, test-debugger, and testing-auditor, which are designed to ingest and process xcresult bundles, failure screenshots, and project source code. \n- Boundary markers: There are no explicit instructions or delimiters defined in the routing logic to warn sub-agents about potentially malicious instructions embedded within the processed test results or code. \n- Capability inventory: The skill description and routing logic mention capabilities such as executing shell commands (xcodebuild, xcresulttool), parsing files, and automatically modifying source code to fix test failures. \n- Sanitization: The skill does not specify any sanitization or validation steps for the external data before it is processed by the routed agents.

axiom-ios-testing