@baize-ai/core@0.2.0
Baize (白泽) — autonomous AI agent infrastructure
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is an automatic lifecycle mutation of AI-agent configuration, but it is gated on an existing Baize installation and targets the package-owned Baize project. The source supports a lifecycle-risk warning rather than evidence of credential theft, remote payload execution, or malware.
Decision evidence
public snapshot- The manifest automatically runs a postinstall script.
- When an initialized Baize directory exists, postinstall always launches the settings and hook synchronizer.
- The synchronizer activates Codex changes if Baize uses Codex or any existing global Codex configuration or authentication file exists.
- It rewrites global Codex configuration to trust the Baize project and enable hooks, then installs SessionStart commands.
- The postinstall exits in CI and does nothing unless the Baize directory already contains .claude.
- The installed hooks target the Baize project and preserve non-core hook groups.
Source & flagged code
33 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 unpkgPackage source references child process execution.
cli/lib/sync-settings-hooks.jsView on unpkg · L21Manifest-reachable source sends a prompted API credential to a fixed unofficial gateway and persists the redirection.
cli/commands/init.jsView on unpkgSource gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.
cli/commands/init.jsView on unpkg · L9Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.
cli/commands/init.jsView on unpkgA manifest entrypoint or package-local install chain reaches persistence behavior.
cli/commands/init.jsView on unpkg · L9Source file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/commands/init.jsView on unpkgPackage source references dynamic require/import behavior.
cli/commands/init.jsView on unpkg · L62Source writes installer persistence such as shell profile or service configuration.
cli/commands/init.jsView on unpkg · L9Package source references weak cryptographic algorithms.
skills/web-console/node_modules/cookie-signature/index.jsView on unpkg · L12A single source file combines environment access, network access, and code or shell execution; review context before blocking.
cli/commands/shell.jsView on unpkg · L8Source creates an unconsented AI-agent control surface through install-time mutation or a default unauthenticated remote skill channel.
scripts/postinstall.jsView on unpkg · L17Runtime or CLI source writes behavior-bearing configuration into a user or project AI-agent control surface.
skills/create-skill/scripts/init-skill.jsView on unpkgPackage source invokes a package manager install command at runtime.
cli/lib/upgrade.jsView on unpkg · L509Package ships native binary artifacts.
skills/comm-bridge/node_modules/better-sqlite3/build/Release/better_sqlite3.nodeView on unpkgPackage ships non-JavaScript build or shell helper files.
docker/entrypoint.shView on unpkgPackage ships high-entropy non-source blobs.
.npm-cache/_cacache/content-v2/sha512/61/51/888f691a98b493c70e8db198e80717d2c2c9f4c9c75eb26738a7e436d5ce733ee675a65f8d7f155dc4fb5d1ef98d54e43a5d2606e0052dcadfc58bb0f5e9View on unpkgPackage ships compressed or archive-like blobs.
.npm-cache/_cacache/content-v2/sha512/61/51/888f691a98b493c70e8db198e80717d2c2c9f4c9c75eb26738a7e436d5ce733ee675a65f8d7f155dc4fb5d1ef98d54e43a5d2606e0052dcadfc58bb0f5e9View on unpkgPackage hides binary, compressed, or executable-looking payloads in test/fixture/hidden paths.
test/integration/runtime/run.shView on unpkgPackage ships a nested archive or MCP bundle that was inventoried but not recursively analyzed.
skills/comm-bridge/node_modules/tar-fs/test/fixtures/invalid.tarView on unpkgA bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
cli/lib/runtime-setup.js#virtual:normalized:round1View on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/commands/doctor.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/runtime-setup.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/commands/runtime.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/commands/self-uninstall.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/commands/service.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/caddy.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/claude-eval.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/download.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/github.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/runtime/claude.jsView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
cli/lib/self-upgrade.jsView on unpkgFindings
Affected versions and remediation
This report applies to @baize-ai/core@0.2.0.
See all versions of @baize-ai/core for other recorded verdicts.
- Review the evidence and your use of @baize-ai/core@0.2.0 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.