Security audit
Security checks for vulnerabilities and agentic risk
This integration needs review because setup automatically executes unverified installation code and chat requests are incorrectly classified as read-only.
Review setup before installing. Install a verified CLI separately and require approval for further installation changes. Check or disable its documented automatic agent-skill synchronization. Treat chat generation as sending selected content through OOMOL to GroqCloud, with possible usage charges. The remote installers and hosted connector implementation were not verified; the bundled artifact provides no evidence of intentional malicious behavior.
The supplied command accesses an Oomol installation endpoint and executes its response locally. Nothing in the supplied chunk connects this installation to GroqCloud or implements the declared data operations. The remote script's contents are not supplied, so its further effects cannot be determined.
Piping an HTTPS response directly into bash executes mutable remote code without version pinning, independent integrity verification, or inspection. This is a conventional installation pattern rather than evidence of malicious intent, but a compromised publisher or distribution endpoint could gain the agent process's permissions and access local credentials or files; it overlaps the fallback-installation finding.
oo: command not found — install the oo CLI (other platforms: https://cli.oomol.com/install-guide.md):
curl -fsSL https://cli.oomol.com/install.sh | bash # macOS / Linux
The skill leaves create_chat_completion untagged while declaring all untagged actions to be safe reads. This incorrectly exempts an inference operation that transmits prompts and can consume paid quota from the skill's payload-and-effect confirmation requirement, potentially enabling unintended submissions or charges.
The chat-completion action is untagged, so the safety instructions classify it as a read that is safe to run directly. Combined with the action-execution workflow, this permits sending user prompts and included context to GroqCloud through OOMOL without warning the user about that disclosure.
The fallback setup directs the agent to execute downloaded shell or PowerShell scripts without a review or explicit installation-approval step. If the distribution endpoint or delivery pipeline is compromised, this can execute attacker-controlled code with the agent's local privileges; the missing-CLI condition limits when this happens but does not protect execution.
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
- **`OOMOL_INSUFFICIENT_CREDIT`** — OOMOL billing stop. Recharge at `https://console.oomol.com/billing/token-recharge` before retrying.
- **HTTP 402 without `OOMOL_INSUFFICIENT_CREDIT`** — the third-party account itself has no active plan or credit. Report the provider error message as is; do not ask the user to recharge OOMOL.
## Resources