@softure-ai/skills@0.2.0
SOFTURE agent skills: a spec-driven delivery workflow (shape → prd → roadmap → plan → implement → review) plus parallel worktree orchestration
AI Security Review
scanned 11d ago · by lpm-firewall-aiThe automatic postinstall setup modifies project agent instructions and installs skills with agent and shell capabilities, warranting a warning under the agent extension lifecycle policy. The inspected code uses managed package-owned content and showed no concrete malicious attack behavior.
Decision evidence
public snapshot- The package runs the installer from its postinstall hook, which targets the directory where npm install was run by default.
- The installer copies package-owned skills into the consumer project's .claude/skills directory and adds managed rules to AGENTS.md or CLAUDE.md without a separate setup prompt.
- A copied skill can use Bash and Agent tools and describes autonomously carrying out a workflow, so the installed agent capability is behavior-bearing.
- The installer skips an existing skill folder unless its manifest says the package manages it.
- Rule sections are configurable, and the installer marks inserted content as managed; it records installed paths in a manifest.
- Uninstall removes skills and managed rule blocks using the recorded manifest. No network endpoint or credential handling was identified in the inspected installer files.
Source & flagged code
8 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 unpkgThe package runs the installer from its postinstall hook, which targets the directory where npm install was run by default.
package.jsonView on unpkg · L22Findings
Affected versions and remediation
This report applies to @softure-ai/skills@0.2.0.
See all versions of @softure-ai/skills for other recorded verdicts.
- Review the evidence and your use of @softure-ai/skills@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.