game-accessibility
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it is designed to ingest and process the content of a target game project's codebase.
- Ingestion points: The agent is instructed to "Investigate the codebase thoroughly" and processes input provided through
$ARGUMENTS(SKILL.md). - Boundary markers: The skill does not implement delimiters or specific instructions to ignore embedded commands within the audited files, which could lead the agent to follow malicious instructions hidden in the project code.
- Capability inventory: The agent has the capability to read arbitrary files within the project and write execution results to a telemetry file in the user's home directory (SKILL.md).
- Sanitization: There is no mention of sanitizing or filtering the content ingested from the external codebase before processing.
- [DATA_EXPOSURE]: The skill contains instructions to automatically write execution metadata, including success outcomes and suggestions, to a hidden platform directory at
~/.claude/projects/(SKILL.md). This establishes a pattern of automated file system modifications based on execution results.
Audit Metadata