@uipath/conversational-tool
Handle conversations with deployed UiPath conversational processes.
This report shows the evidence for @uipath/conversational-tool@1.205.0. 0 of 7 known versions are blocked or flagged. See all versions.
AI Security Review
scanned 1d ago · by lpm-firewall-aiThe package appears to implement a UiPath conversation CLI, and its service requests use the configured service origin. However, it also sends command parameters and error details through runtime telemetry, with no disclosure or opt-out established in the inspected source, so this warrants a warning.
Decision evidence
public snapshot- The CLI tracks command invocations and includes command parameters and error messages in telemetry properties.
- The bundled telemetry provider sends event properties to an Application Insights ingestion endpoint.
- The default telemetry connection string is Base64-encoded, though it decodes to an Azure Application Insights connection string.
- The manifest has no lifecycle scripts; its entry points expose a CLI and command registration for UiPath conversational processes.
- Citation downloads reject URLs outside the configured service origin before attaching the bearer token.
Source & flagged code
5 flaggedSource decodes a Base64-obscured HTTP endpoint at runtime.
dist/tool-5vfakw75.jsView on unpkg · L4631The CLI tracks command invocations and includes command parameters and error messages in telemetry properties.
dist/tool-5vfakw75.jsView on unpkg · L6142The default telemetry connection string is Base64-encoded, though it decodes to an Azure Application Insights connection string.
Findings
Affected versions and remediation
This report applies to @uipath/conversational-tool@1.205.0.
The versions list below shows the verdict for each recorded version.
- Review the evidence and your use of @uipath/conversational-tool@1.205.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.