Skillv1.3.1

ClawScan security

Agent-Powered P2P Energy Trading for Prosumer Microgrids · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 15, 2026, 7:10 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The guide's declared requirements (GreenHelix gateway + Stripe keys) broadly match a trading/settlement guide, but there are provenance inconsistencies and sensitive privileges requested that a user should not hand over without review.
Guidance: This guide can be useful, but treat it like runnable instructions from an unknown author: 1) Do NOT paste production Stripe or gateway keys into a skill you don't fully trust—use Stripe test keys or restricted API keys scoped only to PaymentIntent creation; 2) Verify what 'GREENHELIX_API_KEY' actually grants and prefer sandbox/test credentials; 3) Review all code examples in the SKILL.md before executing anything: look for API calls that dispatch device commands or create/confirm payments; 4) Because the skill can guide an agent to control hardware and move money, run examples in an isolated test environment and behind feature flags; 5) Prefer local/manual execution of examples rather than giving the agent autonomous access to real credentials; 6) Assess provenance: there's no homepage or known publisher—consider contacting the author or finding an alternate vetted source; 7) If you decide to use it, rotate keys afterward and monitor logs for unexpected API activity.

Review Dimensions

Purpose & Capability: noteThe skill's name and description (P2P energy trading, escrow, settlement) reasonably explain why it would need a GreenHelix gateway API key and a Stripe API key for payments. However the SKILL.md repeatedly advertises a GreenHelix 'sandbox' that requires no API key for getting started while still declaring GREENHELIX_API_KEY as the primary credential—this is a minor internal inconsistency. The package has no homepage or known source, which reduces provenance and raises a question about trust but does not by itself contradict the stated purpose.
Instruction Scope: concernThis is an instruction-only guide that includes examples for prosumer registration, smart meter integration, dispatching battery commands, escrow and Stripe settlement. Those actions can control physical devices and move money. The instructions (as advertised) therefore require careful review: the skill could guide an agent to invoke APIs that trigger device control or financial transactions if the user supplies production credentials. The SKILL.md does not appear to limit keys to test scopes (despite mentioning sandbox) or explicitly instruct safe scoping/verification before executing code; that grants broad discretion to follow harmful actions if keys are provided.
Install Mechanism: okNo install spec and no code files — the skill is instruction-only. That minimizes supply-chain risk because nothing is downloaded or executed by default.
Credentials: concernThe skill requests two high-sensitivity secrets: GREENHELIX_API_KEY (primary) and STRIPE_API_KEY. These are proportionate to trading and settlement functionality, but the SKILL.md's sandbox claim (no API key required) conflicts with declaring GREENHELIX_API_KEY as required. Stripe keys are particularly sensitive; the skill name/description justify a payment integration, but the user should only provide test/restricted keys and confirm required scopes. No secondary supporting credentials are requested (good), but the lack of provenance and the potential to execute device control + payments means the credential request deserves caution.
Persistence & Privilege: okThe skill does not request permanent inclusion (always:false), and there is no install step that would modify other skills or system-wide settings. It is user-invocable and can run autonomously (the platform default), which is expected for agent skills; this alone is not flagged.