software-ios-runtime-debugging

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references the XcodeBuildMCP tool and its associated documentation hosted by Sentry on GitHub. Sentry is a well-known technology vendor, and these references are provided for legitimate diagnostic and automation purposes.
  • [COMMAND_EXECUTION]: The skill directs the agent to use standard iOS development commands including xcodebuild, simctl, codesign, and security. These commands are used as intended for verifying build success, managing simulators, and inspecting bundle entitlements.
  • [INDIRECT_PROMPT_INJECTION]: The skill involves reading and interpreting simulator logs and UI hierarchy data. While this presents a theoretical surface for indirect prompt injection, it is the primary purpose of a debugging skill and no exploitable chain was found.
  • Ingestion points: Simulator logs via simctl log stream, UI hierarchy screenshots, and built .app bundle metadata.
  • Boundary markers: No explicit instruction to wrap external data in delimiters is provided.
  • Capability inventory: The agent has capabilities for filesystem modification (rm -rf on DerivedData), bundle signing verification, and simulator process control.
  • Sanitization: The skill does not describe specific sanitization steps for ingested log data before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 09:10 PM
Security Audit — agent-trust-hub — software-ios-runtime-debugging