@trypattern/sdk@0.1.14
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe reachable starter-pull output encourages an assistant to publish a change without owner approval, creating a concrete but conditional risk. The inspected source does not establish autonomous publication or agent configuration mutation, so a warning is appropriate.
Decision evidence
public snapshot- `app pull` is an exposed CLI command that downloads and installs a site.
- For starter sites, generated onboarding text tells an assistant to publish a first change without asking.
- That instruction is returned only when the pulled site is marked as a starter.
- The flagged `.claude` references concern formatting and empty-folder handling, not agent configuration writes.
- The manifest exposes a CLI and declares development scripts; no install lifecycle hook is shown.
- The publish instruction is displayed as CLI output; this source does not itself execute the publish command.
- The inspected agent-related locations do not show global or foreign agent settings being modified.
Source & flagged code
9 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/cli/publishing.jsView on unpkgPackage source references child process execution.
dist/cli/publishing.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @trypattern/sdk@0.1.14.
See all versions of @trypattern/sdk for other recorded verdicts.
- Review the evidence and your use of @trypattern/sdk@0.1.14 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.