Security audit
Security checks for vulnerabilities and agentic risk
The skill is mostly a normal TypeSafe AI connector, but it under-describes external content evaluation and includes risky CLI installation guidance.
Review this before installing. It can use your OOMOL-connected TypeSafe AI account and may send text or JSON you provide to the external TypeSafe AI connector for evaluation. Avoid submitting secrets or sensitive documents unless you intend that sharing, and install the oo CLI through a verified method rather than blindly running the pipe-to-shell commands.
The safety section states that untagged actions are reads, yet evaluate is untagged even though it processes and sends supplied content to an external connector. This incorrect classification can cause agents to run evaluate without heightened checks, potentially exposing sensitive prompts, documents, or structured data under the false assumption that the action is harmless and read-only.
The skill instructs users to install software via curl ... | bash, which executes a remotely fetched script without prior verification. If the distribution endpoint, transport, or hosting is compromised, this can lead to arbitrary code execution on the user's system.
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 manifest and description constrain the skill to 'searching and reading data', but the documented evaluate action accepts arbitrary user text/JSON and performs analysis. This mismatch can mislead an agent or user into treating the skill as read-only, weakening scrutiny and enabling unintended data processing or transmission to an external service.
The trigger language says to use this skill for 'ANY TypeSafe AI request,' which is overly broad and may cause the agent to invoke it for tasks beyond the user's intent or beyond the safest available method. Overbroad routing increases the chance of unnecessary external data sharing and bypasses more context-appropriate controls.
No suspicious patterns detected.