@echomem/mcp@1.4.70
EchoMem MCP bridge for cross-agent memory, local history import, and recall
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe manifest has no install lifecycle hook, and the flagged credential flow uses the EchoMem API base rather than a hard-coded unrelated collector. Startup maintenance is limited to existing EchoMem-owned agent guidance and lifecycle hooks; analytics is described as restricted to allow-listed product metadata with an opt-out.
Decision evidence
public snapshot- package.json has no install, preinstall, or postinstall script; its scripts are for build, start, development, and checks.
- The API client sends the device token to the configured EchoMem API base, which defaults to echo-mem-chrome.vercel.app.
- Default server startup refreshes EchoMem-owned agent guidance and existing lifecycle-hook command paths; the refresh functions leave files without those existing markers or hooks alone.
- The analytics code describes its collected fields as allow-listed and excludes emails, paths, memory text, queries, tokens, and keys.
- Analytics events carry app/version/platform/architecture, installation ID, command, and event data to PostHog; the code supports telemetry opt-out.
- Server startup does write to agent configuration files, but only to refresh EchoMem-owned guidance or existing EchoMem lifecycle hooks.
- The package includes process execution and automatic update code, but the reviewed paths stage this package's own runtime and do not show unrelated code execution.
Source & flagged code
7 flaggedPackage source references child process execution.
dist/durable-entry.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @echomem/mcp@1.4.70.
See all versions of @echomem/mcp for other recorded verdicts.
- Review the evidence and your use of @echomem/mcp@1.4.70 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.