spacecode@2.16.59
SpaceCode - launch cloud dev environments with AI coding agents
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe inspected credential paths are consistent with the package’s cloud-agent features: an explicitly supplied model key is separated from the enabled-by-default SpaceCode proxy path, and stored SpaceCode credentials go to the SpaceCode backend. Scanner similarity and credential-redirection hints were not confirmed as malicious behavior in source.
Decision evidence
public snapshot- package.json exposes the CLI bundles and lists dependencies, with no lifecycle hook.
- A user-supplied Anthropic key is passed to the sandbox; the proxy is configured only when no custom key is supplied and proxy use is enabled.
- The MCP client defaults to SpaceCode’s backend and uses the stored SpaceCode access token for API calls.
- The inspected dependency entrypoints are ordinary SDK, styling, and command-line parsing code.
- Static hints report source similarity and credential redirection, but inspected source shows no unrelated fixed recipient for the user-supplied key.
- The CLI has network and process capabilities for its cloud-session and REPL functions; inspected paths do not establish an attack.
Source & flagged code
12 flagged · loading sourceManifest-reachable source sends a prompted API credential to a fixed unofficial gateway and persists the redirection.
dist/index.jsFindings
Affected versions and remediation
This report applies to spacecode@2.16.59.
See all versions of spacecode for other recorded verdicts.
- Review the evidence and your use of spacecode@2.16.59 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.