Intercept a clawhub install request and run a quick safety check — before anything is written to disk.
Signals included (free version — 3 of 10):
SignalTierWhat it catchesshell+network comboHIGHexec + outbound network in same skill — classic exfil patternhardcoded-external-urlHIGHRaw external URLs embedded in instructionsunverified-publisherMEDIUMPublisher has no verified badge on clawhub.com
Not included (full version — Security Pack):
curl | bash, base64 -d | bash)Get all 10 signals → ClawHub Security Pack
"Install [skill-name] from ClawHub" — runs automatically before install"Is [skill-name] safe to install?" — on-demand scan"clawhub install [skill-name]" — intercepts and scans firsttext
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔍 Pre-Install Scan (Free — 3/10 signals): <skill-name>
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Publisher : <name> [verified / unverified]
Version : <version>
Risk : LOW | MEDIUM | HIGH
Flags:
⚠️ <signal> — <one-line explanation>
Summary: <1–2 sentences>
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
*Full 10-signal scan: https://theagentgordo.gumroad.com/l/clawhub-security-pack*
RatingActionLOWProceed with installMEDIUMWarn user, ask for confirmationHIGHBlock install. Requires --force to override
web_fetch — to retrieve the SKILL.md from ClawHubweb_search — to check publisher standinghttps://clawhub.com/@ordo-tech | Full pack: https://theagentgordo.gumroad.com/l/clawhub-security-pack