composio-br
Warn
Audited by Socket on Nov 4, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill’s broad SaaS automation purpose matches many capabilities, but its actual trust model is inconsistent: it routes actions and identity through a custom gateway rather than official Composio/public vendor endpoints, and it later instructs direct use of COMPOSIO_API_KEY from a workspace .env despite claiming the agent never handles that key. This third-party gateway plus credential-handling inconsistency and high-impact autonomous actions make the skill high risk, though not confirmed malware.
Confidence: 91%Severity: 84%
Audit Metadata