whiyh-gpt-hill-climb

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow reads GPT game telemetry and, when aggregates justify it, loads misses/model-results transcripts via npm run telemetry ... --include-transcript; those transcripts are outsider-authored free text (player messages/answers) that the operating user did not choose to introduce, and the agent then ingests that prose into the LLM context for analysis.