Back to skill
Skillv1.0.0

ClawScan security

Sportfolio.Market · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 14, 2026, 2:17 AM
Verdict
Benign
Confidence
high
Model
gpt-5-mini
Summary
The skill is an instruction-only helper for interacting with the Sportfolio MCP endpoint or a repo-local CLI; its requested actions, endpoints, and authentication model are coherent with that purpose and it does not request unrelated credentials or install anything.
Guidance
This skill is coherent and low-risk in structure, but before using it: (1) only provide a user-scoped Sportfolio token (not admin credentials) and verify its scope; (2) avoid pasting tokens into shared shell history or logs; (3) confirm the MCP URL (https://www.sportfolio.market) is legitimate before connecting; (4) be aware the skill references a local endpoint (127.0.0.1:5000) for development — connecting to loopback can reach local services, so only use that if you trust the local server; and (5) rely on the skill's explicit confirm step before letting it execute staged trades or other mutations.

Review Dimensions

Purpose & Capability
okThe name and description match the SKILL.md: it documents MCP usage, repo-local CLI commands, and the use of a user-scoped API token. There are no unrelated environment variables, binaries, or installs required.
Instruction Scope
noteSKILL.md stays on-topic: it tells the agent how to authenticate (Bearer token), which MCP endpoints to use, and which CLI commands are available, and it explicitly excludes admin/billing flows and requires explicit user confirmation for staged mutations. Note: it references a local endpoint (http://127.0.0.1:5000) for dev use — that legitimately supports local testing but also means an agent following these instructions could attempt connections to loopback services if permitted by the runtime.
Install Mechanism
okInstruction-only skill with no install spec and no code files, which minimizes on-disk footprint and install-time risks.
Credentials
okThe skill expects use of a user API token (described as coming from 'Profile -> CLI Access') and does not require unrelated secrets, config paths, or multiple credentials. It warns not to treat tokens as admin keys.
Persistence & Privilege
okalways:false and no special persistence requested. The skill allows autonomous invocation by default (platform default) but the instructions explicitly require confirmation-gated actions; there are no indications the skill attempts to elevate its privileges or persist state across agents.