Warn by default; block when configured. This changes an agent control surface during installation and may affect how agents behave in the consuming project.
AI security review
scanned · by lpm-firewall-ai
The postinstall hook automatically writes the bundled agent skill into the consuming project’s .agents/skills directory. Under the review rules, install-time writes to agent skills outside the package without consent are malicious.
Package installation through npm, unless the install is global or NEWAPI_CLI_SKIP_SKILL=1 is set.
Impact
This changes an agent control surface during installation and may affect how agents behave in the consuming project.
Mechanism
The hook copies the bundled skill into the project’s `.agents/skills` directory without an explicit setup command.
Policy narrative
During npm postinstall, the package copies its bundled skill into the consuming project’s `.agents/skills` directory. This installs agent-related content outside the package without an explicit user setup command, creating an unconsented agent control surface.
AI identified this as a dangerous AI-agent capability at 98.0% confidence with low false-positive risk.
Evidence for policy risk
The manifest runs `node lib/install.js` as an npm postinstall hook.
Postinstall copies the bundled skill into the consuming project’s `.agents/skills` directory automatically; it only skips this for global installs or when opted out.
The installer calls `installSkill()` before handling the optional binary download skip.
Evidence against
The flagged metadata address appears in a test that expects the image downloader to reject it.
The image downloader requires HTTPS and validates resolved addresses before connecting.
The installer downloads the CLI binary from the configured GitHub release base and verifies its SHA-256 checksum.
Avoid installing @aswless_854771076/newapi-cli@2026.11.12-3. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, check your AI agents' configuration and instruction files (the files this report lists) for changes you did not make, and remove them. This finding does not by itself establish malware intent.