@agent-native/skills@0.3.52
Install BuilderIO skills for coding agents.
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe package's default CLI telemetry sends exception stacks to the publisher's analytics service, which can expose local paths and conflicts with the source's stated privacy limits. This is a runtime telemetry concern, so it warrants a warning rather than a publish block.
Static reason
No blocking static signals were detected.
Trigger
Running a non-help CLI command with telemetry enabled; exception details are sent when the command fails.
Impact
Error messages and stack traces may disclose local paths to the analytics service, despite the source's stated privacy limits.
Mechanism
The CLI captures error messages and exception stacks, adds them to telemetry events, and POSTs those events to the first-party analytics endpoint. Opt-out environment variables are available.
AI rationale
The package's default CLI telemetry sends exception stacks to the publisher's analytics service, which can expose local paths and conflicts with the source's stated privacy limits. This is a runtime telemetry concern, so it warrants a warning rather than a publish block.
Evidence
dist/index.jsdist/telemetry.js~/.agent-native/installation-id
Network endpoints1
analytics.agent-native.com/track
Decision evidence
public snapshotAI called this Suspicious at 92.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The CLI telemetry defaults to POSTing events to https://analytics.agent-native.com/track.
- When a CLI command fails, it reports the error message and captures the exception stack; the telemetry code includes that stack in event properties and POSTs the event. Stack traces can contain local paths.
- Telemetry can be disabled with DO_NOT_TRACK=1 or AGENT_NATIVE_TELEMETRY_DISABLED=1.
- The source says telemetry never reports file paths, but exception stacks are included in captured events.
Evidence against
- The telemetry runs through CLI command handling, not an npm lifecycle hook or import-time action.
- The package is a skills installer, and its stated telemetry fields include skill names, client IDs, scope, counts, platform, and random install and run IDs.
- The package declares no install, preinstall, or postinstall scripts.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
Source & flagged code
5 flaggeddist/index.jsView file
1import { spawnSync } from "node:child_process";
L2: import { randomUUID } from "node:crypto";
High
dist/index.js#virtual:normalized:round1View file
32--force Overwrite a different existing PR Visual Recap workflow
L33: --no-mcp Install skill files only; skip registering the app"s MCP server\n --no-connect Register MCP where possible but skip inline browser/devic...
L34: return file;
High
Same File Env Network Execution
A single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/index.js#virtual:normalized:round1View on unpkg · L32•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1
Findings
4 High4 Medium4 Low
HighChild Processdist/index.js
HighShell
HighSame File Env Network Executiondist/index.js#virtual:normalized:round1
HighSemantic Analysis Limiteddist/index.js#virtual:normalized:round1
MediumNetwork
MediumEnvironment Vars
MediumAi Review Evidencedist/telemetry.js
MediumAi Review Evidencedist/telemetry.js
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @agent-native/skills@0.3.52.
See all versions of @agent-native/skills for other recorded verdicts.
- Review the evidence and your use of @agent-native/skills@0.3.52 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.