owlmem@0.4.1
OwlMem command line: end-to-end encrypted vault for your AI agents, and a local MCP server.
AI Security Review
scanned 11d ago · by lpm-firewall-aiThe scanner’s command and network signals correspond to explicit vault features: authenticated API calls and a caller-supplied command runner. The inspected source shows no install-time execution or independent exfiltration path, so the package is clean.
Static reason
One or more suspicious static signals were detected.
Trigger
A user invokes a CLI command or calls the local MCP vault tool.
Impact
No unauthorized collection or transfer was established from the inspected source.
Mechanism
The program sends bearer-authenticated API requests and can pass requested vault values to a supplied command; it masks matching values in returned output.
AI rationale
The scanner’s command and network signals correspond to explicit vault features: authenticated API calls and a caller-supplied command runner. The inspected source shows no install-time execution or independent exfiltration path, so the package is clean.
Evidence
owlmem.mjs
Network endpoints1
owlmem.com
Decision evidence
public snapshotAI called this Clean at 97.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes one CLI entrypoint and defines no lifecycle scripts.
- Requests send a bearer token to the configured API base, which defaults to owlmem.com.
- Sensitive vault operations use an approval-required flow that waits for the owner’s response.
- The MCP command runner uses a caller-supplied command and vault references; returned output is masked.
Evidence against
- The MCP runner can execute shell commands with vault values in their environment, a capability that could be misused if invoked with an untrusted command.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetwork
HighEntropyStringsUrlStrings
Source & flagged code
7 flaggedowlmem.mjsView file
27import { join, dirname } from "node:path";
L28: import { spawn } from "node:child_process";
L29:
High
27import { join, dirname } from "node:path";
L28: import { spawn } from "node:child_process";
L29:
L30: export const VERSION = "0.4.1";
L31: const BASE = (process.env.OWLMEM_URL || "https://owlmem.com").replace(/\/$/, "");
L32: const DIR = process.env.OWLMEM_HOME || join(process.env.XDG_CONFIG_HOME || join(homedir(), ".config"), "owlmem");
High
Same File Env Network Execution
A single source file combines environment access, network access, and code or shell execution; review context before blocking.
owlmem.mjsView on unpkg · L2721// Reading, writing, deleting and exporting need the space owner's approval in the console; this program waits
L22: // for it (up to 5 minutes). Secrets go to stdout only; messages go to stderr.
L23:
...
L27: import { join, dirname } from "node:path";
L28: import { spawn } from "node:child_process";
L29:
L30: export const VERSION = "0.4.1";
L31: const BASE = (process.env.OWLMEM_URL || "https://owlmem.com").replace(/\/$/, "");
L32: const DIR = process.env.OWLMEM_HOME || join(process.env.XDG_CONFIG_HOME || join(homedir(), ".config"), "owlmem");
L33: const enc = new TextEncoder(), dec = new TextDecoder();
L34: const say = (...a) => process.stderr.write(a.join(" ") + "\n");
L35: class Fail extends Error {}
Findings
5 High2 Medium5 Low
HighChild Processowlmem.mjs
HighSame File Env Network Executionowlmem.mjs
HighCommand Output Exfiltrationowlmem.mjs
HighTrigger Reachable Command Output Exfiltrationowlmem.mjs
HighSemantic Analysis Limitedowlmem.mjs
MediumNetwork
MediumEnvironment Vars
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowSuspicious Lifecycle Evidencepackage.json
LowAi Review Evidenceowlmem.mjs
Affected versions and remediation
This report applies to owlmem@0.4.1.
See all versions of owlmem for other recorded verdicts.
- Review the evidence and your use of owlmem@0.4.1 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.