Skillv1.1.2

ClawScan security

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

Scanner verdict

SuspiciousFeb 11, 2026, 8:31 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill largely does what it says (fetches a Fear & Greed index from a PRISM endpoint and prints it), but its package and instructions claim drop-in React/HTML components and other assets that are not actually included — and it relies on third‑party CDN/APIs; this mismatch and external dependency warrant caution.
Guidance: This skill will fetch sentiment from an external PRISM endpoint and print or return JSON — it does not include the promised React/JS widget files in the package and instead references external CDNs. Before installing/using: 1) Verify the external domains (strykr-prism.up.railway.app, cdn.strykr.com, widgets.strykr.com) and the repository/author to confirm you trust those hosts; 2) Be aware the included script requires curl and jq on PATH (not declared); 3) If you expected local, drop-in components, know that assets will be loaded from third-party CDNs (which can change behavior or collect usage data); 4) Inspect or run the shell script in a sandbox before allowing an agent to execute it; and 5) If you need stronger assurance, ask the publisher for the actual React/JS source and a canonical homepage or release host (GitHub release or official domain).

Review Dimensions

Purpose & Capability: concernThe description advertises drop-in React/HTML components and a distributable widget, but the bundle does not include any React/JS component files or widget code — only a shell script (scripts/fear-greed.sh), SKILL.md, and skill.json. The SKILL.md shows embed snippets that load assets from cdn.strykr.com and widgets.strykr.com, so the package appears to rely on external CDN-hosted assets rather than providing them. That mismatch between claimed deliverables and actual contents is incoherent and should be verified with the author.
Instruction Scope: noteRuntime instructions are narrowly scoped: they call the included shell script which performs a single HTTP GET to PRISM_URL and formats the result. The SKILL.md references external endpoints (strykr-prism.up.railway.app, cdn.strykr.com, widgets.strykr.com). The script only uses PRISM_URL and does not read other files or environment variables, nor does it exfiltrate local data — however it assumes availability of curl and jq (not declared in registry).
Install Mechanism: okThere is no install specification (instruction-only) and no downloads or archive extraction. The only executable is the included shell script. This lowers install-time risk.
Credentials: okNo credentials or secret environment variables are required. The only configurable variable is PRISM_URL (defaulting to https://strykr-prism.up.railway.app). That is proportionate for a widget that queries a public API.
Persistence & Privilege: okThe skill does not request always:true or modify other skills or system configuration. It is user-invocable and can be invoked autonomously (platform defaults) but has no elevated persistence or privileges.