registry / @system-callifly/widget / 0.1.0-beta.7
@system-callifly/widget@0.1.0-beta.7
Callifly's embeddable voice/chat assistant widget — a framework-agnostic Web Component, usable via a <script> tag or npm install.
AI Security Review
scanned 4d ago · by lpm-firewall-aiThe package manifest has no install lifecycle hooks, and the inspected network and dynamic-code behavior aligns with the widget’s stated chat and voice-call purpose. The scanner’s command-execution and exfiltration leads were not confirmed in the inspected entry bundle.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
Widget use; Daily call functionality loads its call-machine bundle when needed.
Impact
The inspected source showed no unrelated command execution or environment-data exfiltration. Runtime behavior includes sending chat data to Vapi and loading Daily call code.
Mechanism
Chat requests go to api.vapi.ai using a configured bearer token. Daily call code is fetched from c.daily.co and executed for call functionality.
AI rationale
The package manifest has no install lifecycle hooks, and the inspected network and dynamic-code behavior aligns with the widget’s stated chat and voice-call purpose. The scanner’s command-execution and exfiltration leads were not confirmed in the inspected entry bundle.
Network endpoints3
api.vapi.aic.daily.cogs.daily.co
Decision evidence
public snapshotAI called this Clean at 92.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest points consumers to the widget’s UMD and ES module builds; its scripts are development, build, test and lint commands, with no install lifecycle hook.
- The chat client sends requests to api.vapi.ai using a configured bearer token, consistent with the widget’s stated voice/chat purpose.
- The Daily call SDK fetches its versioned call-machine bundle from c.daily.co and executes that bundle as part of loading call functionality.
Evidence against
- The UMD build includes dynamic execution of a fetched bundle, so the widget relies on remote code from its Daily call-service host.
- The package depends on browser-focused UI and document-processing libraries, matching the published embeddable widget purpose.
Behavioral surface
ChildProcessDynamicRequireEnvironmentVarsEvalFilesystemNetwork
HighEntropyStringsMinifiedUrlStrings
Source & flagged code
8 flaggeddist/pdf-RgrI1mmZ.jsView file
1715}
L1716: exec(t, e) {
L1717: var c;
High
Child Process
Package source references child process execution.
dist/pdf-RgrI1mmZ.jsView on unpkg · L17156717_createCanvas(t, e) {
L6718: return process.getBuiltinModule("module").createRequire(import.meta.url)("@napi-rs/canvas").createCanvas(t, e);
L6719: }
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/pdf-RgrI1mmZ.jsView on unpkg · L6717dist/index-CAn7n2eC.jsView file
490try {
L491: a = new Function(o, "_", i);
L492: } catch (h) {
Low
Eval
Package source references a known benign dynamic code generation pattern.
dist/index-CAn7n2eC.jsView on unpkg · L490Findings
1 Critical5 High4 Medium6 Low
CriticalPrevious Version Dangerous Deltadist/callifly-widget.umd.cjs
HighChild Processdist/pdf-RgrI1mmZ.js
HighSame File Env Network Executiondist/callifly-widget.umd.cjs
HighCommand Output Exfiltrationdist/callifly-widget.umd.cjs
HighTrigger Reachable Command Output Exfiltrationdist/callifly-widget.umd.cjs
HighSemantic Analysis Limiteddist/callifly-widget.umd.cjs
MediumDynamic Requiredist/pdf-RgrI1mmZ.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowScripts Present
LowEvaldist/index-CAn7n2eC.js
LowFilesystem
Affected versions and remediation
This report applies to @system-callifly/widget@0.1.0-beta.7.
See all versions of @system-callifly/widget for other recorded verdicts.
- Review the evidence and your use of @system-callifly/widget@0.1.0-beta.7 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.