Skillv1.0.2

ClawScan security

The Fed Agent · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 17, 2026, 3:48 PM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's code, runtime instructions, and included reference files are coherent with a macroeconomic / Fed-analysis purpose and do not request unrelated credentials or install arbitrary third‑party code.
Guidance: This skill appears internally consistent for macroeconomic analysis: it fetches a news URL, uses the included frameworks, and outputs trading/policy analysis. Before installing, consider: (1) this will access external news URLs (network access) — only provide URLs you trust; (2) it produces trading recommendations (financial advice risk) — verify outputs before acting; (3) if you plan to run the included scripts locally, you'll need Node.js and should inspect/execute them in a sandbox first — the code is currently placeholder and does not perform hidden network exfiltration; (4) autonomous invocation is allowed by default — if you don't want the agent to run this skill without explicit permission, disable or remove it from autonomous skill sets. Overall there are no red flags in the files provided.

Review Dimensions

Purpose & Capability: okName/description (Fed/macro analysis) match the content: SKILL.md, references, and scripts focus on monetary policy, inflation, and trading implications. Minor note: a Node script (scripts/analyze.js) is included but the skill declares no required binaries — this is plausible (the SKILL.md expects platform web_fetch rather than running the script), but it's a small inconsistency to be aware of if you plan to execute files locally.
Instruction Scope: okSKILL.md restricts runtime activity to fetching the provided news URL (via web_fetch), reading bundled framework references, applying analytical frameworks, and producing a 4‑section output. It does not instruct reading unrelated files, environment variables, or transmitting data to third‑party endpoints beyond the news URL.
Install Mechanism: okNo install spec (instruction-only) and no downloads or external install steps. Included code is local and non-obfuscated placeholder logic; nothing writes/extracts archives or pulls code from untrusted URLs.
Credentials: okThe skill requests no environment variables, credentials, or config paths — appropriate for a read‑only analytical agent that fetches public news articles and uses bundled references.
Persistence & Privilege: okalways is false and the skill does not request persistent system privileges or modify other skills' configuration. Autonomous invocation is allowed (platform default) but not combined with any broad credential access.