Skillv2.1.1

ClawScan security

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

Scanner verdict

BenignMar 10, 2026, 7:32 PM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's files, instructions, and requirements are consistent with a CFO/advisory modeling tool — no unexpected network calls, credentials, or install steps were found.
Guidance: This skill appears coherent and self-contained: it provides reference docs and three standard Python scripts for runway, unit economics, and fundraising modeling, and it does not request credentials or perform network access in the visible code. Before running: (1) review the omitted unit_economics_analyzer.py to confirm it also uses only stdlib and has no network calls; (2) run the scripts locally in a controlled environment (they may write CSV output when invoked with --csv); (3) avoid pasting or uploading sensitive production credentials or raw PII into the chat — the skill will ask for financial context to produce useful outputs, so share only what’s necessary; (4) if you need automated, always-on monitoring or integration with bank/finance systems later, expect to require explicit, service-specific credentials — that's an escalation and should be reviewed separately. If any script contains hidden network requests, external downloads, or asks for environment secrets, reclassify as suspicious.

Review Dimensions

Purpose & Capability: okName/description (CFO advisory, runway, unit economics, fundraising) match the included reference docs and three Python scripts (burn rate, unit economics, fundraising model). There are no unrelated requirements (no cloud creds, no unusual binaries).
Instruction Scope: noteSKILL.md instructs the agent/user to run local Python scripts and to surface certain proactive triggers when company context indicates them. The instructions do not tell the agent to read system config files, environment secrets, or contact external endpoints. Note: the 'proactive triggers' language implies the skill will examine whatever company context the user supplies (or the agent is given) and may prompt for or summarize sensitive financial data — but the skill itself does not contain instructions to access system secrets automatically.
Install Mechanism: okNo install spec is present (instruction-only). The included scripts are pure Python stdlib (no external downloads or package installs), so there is no elevated install-time risk or remote code fetch in the package metadata.
Credentials: okThe skill declares no required environment variables, no primary credential, and the scripts shown use only the Python standard library. There are no requests for unrelated credentials or secrets in SKILL.md or the visible scripts.
Persistence & Privilege: okalways=false and the skill does not request persistent system-wide changes. The SKILL.md's proactive trigger behavior combined with normal autonomous invocation could cause the agent to surface findings proactively, but this is not a privilege escalation or hidden persistence mechanism in the package itself.