iclavue@11.0.35
Clavue: execution-first AI coding CLI with direct repo tools, provider routing, native workflows, MCP integration, and long-session recovery
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package has a concrete unsigned remote native-execution capability, but no source evidence establishes malicious intent or an unconsented install-time attack chain.
Decision evidence
public snapshot- dist/cli.js downloads a platform native pager from clavue.com at runtime.
- The downloaded pager is written under the user config directory, chmodded, and spawned without checksum/signature verification.
- CLAVUE_PAGER_DOWNLOAD_URL can override the downloaded executable URL.
- package.json postinstall only chmods bundled dist/native/clavue-pager files.
- scripts/postinstall-pager.mjs makes no network requests or config changes.
- Provider credentials in dist/openai-responses-adapter.js are used for configured AI-provider requests; no separate exfiltration path found.
- Settings writes target Clavue-owned .clavue paths and are tied to CLI actions.
Source & flagged code
11 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgPackage source references child process execution.
dist/mao-command.jsView on unpkg · L4Findings
Affected versions and remediation
This report applies to iclavue@11.0.35.
See all versions of iclavue for other recorded verdicts.
- Review the evidence and your use of iclavue@11.0.35 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.