Back to skill

Security audit

ClickHouse

Security checks for vulnerabilities and agentic risk

Overview

The ClickHouse workflow has clear safeguards, but setup permits unverified code execution and does not explain the dependency’s broader persistent changes to agent skills.

Review and approve the CLI installation separately, using a verified release. Check its managed-skill behavior and disable startup synchronization with OO_SKILLS_SYNC_DISABLED=1 if those changes are unwanted. Give the ClickHouse connection only the permissions needed and approve exact SQL targets before execution. The bundled instructions were fully inspected, but the remote installers and hosted connector implementation were not verified; no malicious intent was established.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied command installs external software rather than performing ClickHouse operations. Although dependency installation can be a supporting implementation detail, this chunk establishes no connection between the Oomol installer and the declared ClickHouse functionality. The downloaded script's contents and downstream effects are not available for assessment.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

Piping the HTTPS installer response directly into Bash executes mutable remote content without checking its integrity or reviewing it. HTTPS protects transport but not against compromise of the publisher or distribution infrastructure, which could expose local files, credentials, and connected resources accessible to the process. This is the same underlying installer risk identified by SQP-2, not an independent exploit path.

Content

Scanner excerpt · SKILL.md (reported line 59)May include surrounding context.

The setup workflow executes remotely downloaded installer scripts without warning about the system-integrity implications.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

When the CLI is missing, the skill directs the agent to execute remotely fetched shell or PowerShell code without integrity verification or explicit installation approval. A compromised distribution endpoint could execute arbitrary code with the agent's host permissions; limiting this to first-time setup reduces exposure but does not eliminate the risk.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

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.

Content

Scanner excerpt · SKILL.md (reported line 80)May include surrounding context.

md
- **`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