solidcraft

Fail

Audited by Snyk on Nov 17, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.30). The visible skill and reference content document legitimate CAD automation with loopback, token-authenticated control, but the ui.agent_chat parameter allow_insecure_http lets an agent send an API key over unencrypted HTTP without a confirmation dialog, which weakens a protection around the user's credentials beyond what modelling needs.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

  • Potentially malicious external URL detected (medium risk: 0.30). The instructed fetch of upstream skills/ guidance from github.com/tangency-labs/solidcraft (SKILL.md line 82) is an inbound source whose content the agent adopts as instructions, and although tangency-labs appears to be the project's own organization, the skill set also cites an individually owned bherbruck/solidcraft repository as the skills source, so first-party provenance is not clearly established, giving a medium tier.

Issues (2)

E004
CRITICAL

Prompt injection detected in skill instructions.

W012
MEDIUM

Unverifiable external dependency detected (runtime URL that controls agent).

Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 17, 2026, 08:57 AM
Issues
2