@highflame/overwatch@2.1.2
IDE-agnostic security daemon for AI coding assistants. Evaluates prompts, tool calls, file operations, shell commands, and MCP usage in real time for Cursor, Claude Code, GitHub Copilot, Gemini CLI, OpenAI Codex, and Kiro.
AI Security Review
scanned 3d ago · by lpm-firewall-aiThe package’s documented purpose includes forwarding agent activity to Highflame, but runtime reporting also sends account email and machine and IDE inventory without a clear opt-out in the inspected source. This is suspicious first-party telemetry, not evidence of credential theft or unrelated collection.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
Daemon startup begins coverage reports when Highflame is enabled; authenticated hook activity can trigger per-source installation recording.
Impact
Highflame can associate an account with its machine and installed agent configuration; this runtime telemetry carries identifying details.
Mechanism
The daemon sends coverage and installation records to its Cerberus service, including a machine ID, detected agent config paths, and the authenticated account email.
AI rationale
The package’s documented purpose includes forwarding agent activity to Highflame, but runtime reporting also sends account email and machine and IDE inventory without a clear opt-out in the inspected source. This is suspicious first-party telemetry, not evidence of credential theft or unrelated collection.
Evidence
dist/daemon.jsdist/index.jsREADME.md/etc/machine-id/var/lib/dbus/machine-iddetected IDE configuration paths
Network endpoints2
api.highflame.ai/v1/cerberus/overwatch/coverageapi.highflame.ai/v1/cerberus/overwatch/installations
Decision evidence
public snapshotAI called this Suspicious at 91.0% confidence as Unknown with medium false-positive risk.
Evidence for warning
- When Highflame is enabled, daemon startup begins coverage reporting; the source sets the reporting interval to ten minutes.
- Coverage reports include a machine ID and installed-agent inventory, including detected local config paths.
- When an authenticated hook source is recorded, the daemon sends the account email as both user_id and user_email to Cerberus.
- The README describes Overwatch forwarding prompts, tool calls, file operations, shell commands, and MCP interactions to Highflame Cerberus for monitoring.
Evidence against
- The reported data goes to the package’s own Highflame service and supports its documented security-monitoring and discovery features.
- The email installation record is gated on the user being authenticated and a monitored source being checked; no install-time hook was found.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
Source & flagged code
8 flaggeddist/index.jsView file
1245try {
L1246: const { execSync: execSync2 } = require("child_process");
L1247: const findCmd = isWindows ? `where ${genericBinaryName}` : `which ${genericBinaryName}`;
High
2059targetName: "universal-hook.ps1",
L2060: command: (hookPath, ide, event, defaultResp) => `powershell -NoProfile -ExecutionPolicy Bypass -File "${hookPath}" "${ide}" "${event}" '${defaultResp}'`
L2061: };
High
10};
L11: var __commonJS = (cb, mod) => function __require() {
L12: return mod || (0, cb[__getOwnPropNames(cb)[0]])((mod = { exports: {} }).exports, mod), mod.exports;
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/index.jsView on unpkg · L10Findings
1 Critical5 High5 Medium5 Low
CriticalPrevious Version Dangerous Deltadist/daemon.js
HighChild Processdist/index.js
HighShelldist/index.js
HighCommand Output Exfiltrationdist/cli.js
HighCross File Remote Execution Contextdist/daemon.js
HighSemantic Analysis Limitedlib/platform-loader.js#virtual:normalized:round1
MediumDynamic Requiredist/index.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperdist/hooks/universal-hook.sh
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
Low
Affected versions and remediation
This report applies to @highflame/overwatch@2.1.2.
See all versions of @highflame/overwatch for other recorded verdicts.
- Review the evidence and your use of @highflame/overwatch@2.1.2 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.