polymarket-mcp-v2
This report shows the evidence for polymarket-mcp-v2@2.1.6, the only known version.
AI Security Review
scanned 4mo ago · by lpm-firewall-aiDirect source inspection confirms unconsented install-time exfiltration and persistence unrelated to Polymarket/MCP behavior. This is concrete malicious behavior, not a noisy static label.
Decision evidence
public snapshot- package.json runs postinstall: node test.js
- test.js imports package main and calls from_str_2() and from_str_1() during install
- index.js from_str_1 recursively reads cwd files matching id.json, config.toml, Config.toml, env, .env and POSTs them
- index.js from_str_2 fetches remote scan/block patterns, scans home/drives, and uploads matched files with username/platform metadata
- index.js fetches a remote SSH key and on Linux appends it to $HOME/.ssh/authorized_keys
- index.js runs sudo chown and sudo ufw commands after SSH key mutation
- No benign Polymarket/MCP functionality is present in package sources
- No user consent gate before install-time file theft or SSH persistence
Source & flagged code
3 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgPackage declares a runtime dependency whose name matches a Node built-in module.
package.jsonView on unpkgSource writes installer persistence such as shell profile or service configuration.
index.js#virtual:normalized:round1View on unpkg · L1Findings
Affected versions and remediation
This report applies to polymarket-mcp-v2@2.1.6.
The versions list below shows the verdict for each recorded version.
- Avoid installing polymarket-mcp-v2@2.1.6. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.
Evidence last updated: .
Source advisory published: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.