first-tree-staging@0.5.19-staging.1008.1
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 scanner’s command-output/exfiltration claim is not substantiated by the reviewed source; the large shared bundle contains normal CLI, SDK, and telemetry primitives. The package nevertheless has a real first-party agent-extension setup capability that merits a warning under the stated policy.
Decision evidence
public snapshot- dist/cli/index.mjs installs the first-tree-context plugin into Codex or Claude Code via provider CLI commands.
- The plugin has a SessionStart activation bridge that contacts a configured First Tree server.
- package.json postinstall deletes selected Codex/Claude SDK dependency packages during global installs.
- Plugin installation is part of explicit context-integration operations, not the npm lifecycle hook.
- Provider installs are named first-tree-context, validated, journaled, and rollback-aware.
- Postinstall scripts require npm_config_global=true and target enumerated SDK packages only.
- No install-time network, credential harvesting, shell payload, or foreign AI-agent config mutation was found.
Source & flagged code
9 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@0.5.19-staging.1008.1.
See all versions of first-tree-staging for other recorded verdicts.
- Review the evidence and your use of first-tree-staging@0.5.19-staging.1008.1 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.