Security audit
Security checks for vulnerabilities and agentic risk
The Kaggle workflow fits its purpose, but its setup instructions allow unverified remote code execution and under-disclose broader changes to agent skills.
Review the setup before installing. Prefer a verified, pinned CLI installed under your control, and disable automatic synchronization of unrelated agent skills. Understand that OOMOL mediates Kaggle access and may charge for usage. The complete bundled skill was inspected, but remote installers and hosted execution were not verified; no malicious behavior was established.
The supplied command executes an Oomol-hosted installation script. No Kaggle access or data-reading functionality is shown, and no connection establishes this installation as a supporting implementation detail for the declared purpose. The remote script's contents and further effects cannot be determined from this chunk.
The supplied chunk only performs a login operation; it contains no Kaggle resource access, search, or reading behavior. Authentication could support a Kaggle implementation, but the provided code establishes no such connection and instead has authentication as its sole observable purpose.
The curl-to-bash pipeline executes the downloaded response without verifying its contents, version, or provenance beyond HTTPS transport. Compromise of the installer endpoint could turn an otherwise read-oriented Kaggle task into arbitrary local code execution; this is the same underlying installation risk identified by SQP-2, not evidence that the current installer is malicious.
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
Both installation examples execute mutable remote scripts immediately, without an integrity check or an explicit approval step for local code execution. A compromised distribution endpoint could therefore execute arbitrary commands with the invoking user's permissions; the missing-CLI condition reduces exposure but does not remove this supply-chain risk.
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