{"record":{"id":"3ed3e4c57b6515c8","repo":"tinyhumansai/openhuman","slug":"prompt-flagged-for-security-review-and-was-not-pro","errorCode":null,"errorMessage":"Prompt flagged for security review and was not processed.","messagePattern":"Prompt flagged for security review and was not processed\\.","errorType":"validation","errorClass":null,"httpStatus":null,"severity":"error","filePath":"src/openhuman/agent/harness/session/runtime.rs","lineNumber":805,"sourceCode":"    /// This is the primary high-level method for programmatic interaction with the agent.\n    /// It wraps the core `turn` logic with telemetry events (`AgentTurnStarted`,\n    /// `AgentTurnCompleted`) and error sanitization.\n    pub async fn run_single(&mut self, message: &str) -> Result<String> {\n        let guard = enforce_prompt_input(\n            message,\n            PromptEnforcementContext {\n                source: \"agent.runtime.run_single\",\n                request_id: None,\n                user_id: Some(self.event_channel()),\n                session_id: Some(self.event_session_id()),\n            },\n        );\n        if !matches!(guard.action, PromptEnforcementAction::Allow) {\n            let user_message = match guard.action {\n                PromptEnforcementAction::Allow => \"Message accepted.\",\n                PromptEnforcementAction::Blocked => \"Prompt blocked by security policy.\",\n                PromptEnforcementAction::ReviewBlocked => {\n                    \"Prompt flagged for security review and was not processed.\"\n                }\n            };\n            let action_tag = match guard.action {\n                PromptEnforcementAction::Allow => \"allow\",\n                PromptEnforcementAction::Blocked => \"blocked\",\n                PromptEnforcementAction::ReviewBlocked => \"review_blocked\",\n            };\n            crate::core::observability::report_error(\n                user_message,\n                \"agent\",\n                \"prompt_injection_blocked\",\n                &[\n                    (\"session_id\", self.event_session_id()),\n                    (\"channel\", self.event_channel()),\n                    (\"action\", action_tag),\n                ],\n            );\n            BUS.publish(DomainEvent::AgentError {","sourceCodeStart":787,"sourceCodeEnd":823,"githubUrl":"https://github.com/tinyhumansai/openhuman/blob/a221052e0df5b1f7598fceba7329fd1af95d6699/src/openhuman/agent/harness/session/runtime.rs#L787-L823","documentation":"Error \"Prompt flagged for security review and was not processed.\" thrown in tinyhumansai/openhuman.","triggerScenarios":"Thrown at src/openhuman/agent/harness/session/runtime.rs:805 when the library encounters an invalid state.","commonSituations":"See trigger scenarios.","solutions":[],"exampleFix":null,"handlingStrategy":null,"validationCode":null,"typeGuard":null,"tryCatchPattern":null,"preventionTips":[],"tags":[],"backgroundTag":null,"analyzedSha":"a221052e0df5b1f7598fceba7329fd1af95d6699","analyzedAt":"2026-08-16T12:47:06.542Z","schemaVersion":2},"datasetVersion":"2026-08-16T13:17:31.715Z"}