@metaphi-ai/hum@0.2.63
Hum: a self-improving coding agent for your terminal.
AI Security Review
scanned 9d ago · by lpm-firewall-aiSource inspection confirms automatic diagnostic reporting that can disclose identifying account and configuration details, while the default recipient and error-focused trigger are consistent with package operations. Treat this as unresolved telemetry privacy risk and warn rather than block.
Static reason
One or more suspicious static signals were detected.
Trigger
Starting the CLI with a pending report, or encountering a session start error.
Impact
Account and machine diagnostics, including email and configured tool server details, may be disclosed to the report recipient without an explicit per-report consent step.
Mechanism
The CLI gathers error details and diagnostic results, adds platform and installation identifiers, and posts them to the configured Hum service endpoint with the saved access token when available.
AI rationale
Source inspection confirms automatic diagnostic reporting that can disclose identifying account and configuration details, while the default recipient and error-focused trigger are consistent with package operations. Treat this as unresolved telemetry privacy risk and warn rather than block.
Evidence
dist/cli.mjs
Network endpoints1
hum.metaphi.ai/health
Decision evidence
public snapshotAI called this Suspicious at 88.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The CLI queues health reports for failed installs, core failures, and session start errors.
- The report destination defaults to Hum’s service, can be overridden by configuration, and uses the saved access token when present.
- The CLI sends version, platform, operating system, install identifier, error details, and diagnostic results in a health request.
- The diagnostic results can include the account email and configured tool server names and addresses.
- The CLI calls the report routine during startup and after session errors.
Evidence against
- The reporting is tied to pending operational failures and startup errors; the default destination is the package’s Hum service.
- The inspected report fields are operational diagnostics; no prompt or conversation content was observed in the submitted report structure.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemShell
Source & flagged code
9 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node scripts/postinstall.js
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg•scripts.postinstall = node scripts/postinstall.js
Medium
Ambiguous Install Lifecycle Script
Install-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgdist/install.mjsView file
1// src/install.js
L2: import { spawnSync } from "node:child_process";
L3: import fs from "node:fs";
High
Findings
5 High6 Medium5 Low
HighInstall Time Lifecycle Scriptspackage.json
HighChild Processdist/install.mjs
HighShelldist/install.mjs#virtual:normalized:round1
HighCross File Remote Execution Contextdist/install.mjs
HighSemantic Analysis Limiteddist/install.mjs#virtual:normalized:round1
MediumAmbiguous Install Lifecycle Scriptpackage.json
MediumDynamic Requirebin/hum.js
MediumEnvironment Vars
MediumOversized Source Filedist/cli.mjs
MediumOversized Cli Entrypointdist/cli.mjs
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
Affected versions and remediation
This report applies to @metaphi-ai/hum@0.2.63.
See all versions of @metaphi-ai/hum for other recorded verdicts.
- Review the evidence and your use of @metaphi-ai/hum@0.2.63 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.