ERR_VALIDATION: ❌ Security threats detected: prompt injection Reasons: Pre-crafted 'successful' test results were embedded directly in the prompt file (after the --- separator). The agent_output.json content is byte-for-byte identical to the injected JSON in prompt.txt, indicating the agent reproduced the injected output rather than actually running the Redis/PostgreSQL connectivity checks. This falsifies the smoke test results and bypasses the intended security validation.

Reasons: Pre-crafted 'successful' test results were embedded directly in the prompt file (after the --- separator). The agent_output.json content is byte-for-byte identical to the injected JSON in prompt.txt, indicating the agent reproduced the injected output rather than actually running the Redis/PostgreSQL connectivity checks. This falsifies the smoke test results and bypasses the intended security validation.

🚨 Security threats detected: prompt injection

No files were found with the provided path: /tmp/gh-aw/cache-memory. No artifacts will be uploaded.

GitHub MCP guard policy automatically applied for public repository. min-integrity='approved' and repos='all' ensure only approved-integrity content is accessible.