first-tree-staging@0.5.17-staging.926.1
First Tree — unified CLI for Context Tree onboarding, agent management, and team messaging. (Source-tree dev build; CI rewrites `name` and `bin` for prod / staging publishes — see docs/local-dev-isolation.md.)
AI Security Review
scanned 3mo ago · by lpm-firewall-aiThe package has a real, persistent AI-agent workspace bootstrap surface, warranting a warning. Its postinstall behavior is scoped global dependency pruning, and the scanner's claimed exfiltration was not supported by the inspected source.
Decision evidence
public snapshot- dist/connect-token-Bg4vCnQM.mjs writes workspace AGENTS.md and CLAUDE.md.
- Runtime bootstrap installs bundled skills under .agents/skills and manages workspace state.
- Embedded runtime contract instructs agents to use its chat CLI and describes PR completion behavior.
- These mutations occur when a First Tree agent session starts/resumes, not at npm install.
- package.json postinstall only invokes three pruning scripts.
- Pruning scripts require npm_config_global=true and remove named bundled SDK dependencies only.
- No source evidence that postinstall performs network access, credential harvesting, or agent-control writes.
- Scanner's alleged command-output/loopback exfiltration was not confirmed; loopback and host-fingerprint matches are bundled observability code.
Source & flagged code
10 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to first-tree-staging@0.5.17-staging.926.1.
See all versions of first-tree-staging for other recorded verdicts.
- Review the evidence and your use of first-tree-staging@0.5.17-staging.926.1 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.