@echomem/mcp@1.4.38
EchoMem MCP bridge: cloud-first memory tools and the Agent Doctor workspace forensics report (cost ledger + 3D repo city)
AI Security Review
scanned 3mo ago · by lpm-firewall-aiThis is a dangerous, user-invoked AI-agent control and self-update capability, not confirmed malware. The package lacks install-time execution, so it does not meet the blocking threshold.
Decision evidence
public snapshot- dist/setup.js: init invokes setup for all detected agents and enables save hooks by default.
- dist/setup.js writes EchoMem directives into ~/.codex/AGENTS.md and ~/.claude/CLAUDE.md, including preference over other memory providers and automatic saving.
- dist/hud/hooks.js installs Codex/Claude Stop hooks that execute this package's CLI.
- dist/index.js auto-downloads newer package versions after a background registry update check.
- dist/headless-runtime.js runs npm install of the selected newer version into a durable user runtime.
- package.json has no preinstall, install, or postinstall lifecycle hook.
- Agent configuration and hooks are reached only through explicit init/setup/update commands.
- No source evidence of credential harvesting, arbitrary remote code execution, or stealth exfiltration.
- Default network API is package-aligned: echo-mem-chrome.vercel.app.
Source & flagged code
7 flaggedPackage source references dynamic require/import behavior.
dist/headless-runtime.jsView on unpkg · L65Findings
Affected versions and remediation
This report applies to @echomem/mcp@1.4.38.
See all versions of @echomem/mcp for other recorded verdicts.
- Review the evidence and your use of @echomem/mcp@1.4.38 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.