registry / @columenlabs/lumen-ui / 0.1.122
@columenlabs/lumen-ui@0.1.122
Lumen: shared React components (frame, views, widgets, charts, drawers).
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe package is a React component library with no lifecycle scripts, and its reviewed network behavior targets app-provided API routes. Telemetry is an explicitly mounted app hook with a same-origin default endpoint, and account-related actions are exposed through user-driven UI flows.
Static reason
No blocking static signals were detected.
Trigger
The consuming app invokes the relevant client methods or mounts the telemetry hook.
Impact
No unauthorized external data flow or automatic account action was found.
Mechanism
Requests use app-provided API bases; optional telemetry defaults to a same-origin endpoint. Query persistence uses browser localStorage.
AI rationale
The package is a React component library with no lifecycle scripts, and its reviewed network behavior targets app-provided API routes. Telemetry is an explicitly mounted app hook with a same-origin default endpoint, and account-related actions are exposed through user-driven UI flows.
Evidence
dist/telemetry.jsdist/assistant/client.jsdist/people/client.jsdist/query-client.js
Network endpoints3
/api/t (default telemetry endpoint; configurable)/api/assistant (documented default base)/api/people (documented default base)
Decision evidence
public snapshotAI called this Clean at 94.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- Package metadata describes a shared React component library and ends without lifecycle scripts.
- Telemetry starts only when an app calls useLumenTelemetry; its default endpoint is the app-relative /api/t.
- Telemetry is documented as first-party app reporting for screens, errors, failed requests and Web Vitals.
- Assistant and people clients send requests to app-provided base paths rather than hard-coded external hosts.
- Invite acceptance is implemented as a user-facing flow with an explicit Accept action.
Evidence against
- The package includes runtime telemetry capable of sending reports to a configurable endpoint.
- Assistant and people features can make requests using the consuming app's API.
Behavioral surface
FilesystemNetworkWebSocket
HighEntropyStringsUrlStrings
Source & flagged code
7 flaggeddist/assistant/lumen-assistant.js#virtual:normalized:round1View file
•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 5
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/assistant/lumen-assistant.js#virtual:normalized:round1View on unpkgpackage.jsonView file
1{
"name": "@columenlabs/lumen-ui",
"version": "0.1.122",
"description": "Lumen: shared React components (frame, views, widgets, charts, drawers).",
"license": "MIT",
"typ
Low
Suspicious Lifecycle Evidence
Package metadata describes a shared React component library and ends without lifecycle scripts.
package.jsonView on unpkg · L188},
"peerDependencies": {
"@tanstack/react-query": "^5.103.2",
"@tanstack/react-router": "^1.170.39",
"react": "^19.0.0",
Findings
1 High1 Medium9 Low
HighSemantic Analysis Limiteddist/assistant/lumen-assistant.js#virtual:normalized:round1
MediumNetwork
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowSuspicious Lifecycle Evidencepackage.json
LowSuspicious Lifecycle Evidencepackage.json
LowAi Review Evidencedist/telemetry.js
LowAi Review Evidencedist/telemetry.js
LowAi Review Evidencedist/assistant/client.js
LowAi Review Evidencedist/people/client.js
Affected versions and remediation
This report applies to @columenlabs/lumen-ui@0.1.122.
See all versions of @columenlabs/lumen-ui for other recorded verdicts.
- Review the evidence and your use of @columenlabs/lumen-ui@0.1.122 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.