ClawHub
Back to skill

Security audit

Excalidraw MCP

Security checks for vulnerabilities and agentic risk

Overview

The diagram integration is coherent, but its fallback setup runs unverified remote installers without a separate approval step.

Review and approve CLI installation separately, using a specific verified release. Check which additional agent skills it installs, and share only intended diagram content through OOMOL. The publisher documents OO_SKILLS_SYNC_DISABLED=1 to disable startup skill synchronization. [CLI controls](https://github.com/oomol-lab/oo-cli/blob/main/docs/commands.md#environment-variables). The remote installers and backend were not verified; no malicious payload was established.

A.I.G

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

SkillSpector

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 (6)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The actionable risk is the fallback instruction to download and immediately execute installation scripts, extending a diagram-related task into local software installation without an explicit approval or integrity-verification step. The CLI is not unrelated—it is the skill's documented integration mechanism—but compromise of its distribution endpoint could result in code execution with the invoking user's permissions.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding

The supplied code only initiates login; it does not search or read Excalidraw data. Authentication could be a supporting implementation detail, but the supplied chunk establishes no connection between the oo login and Excalidraw MCP. Its visible behavior therefore materially differs from the declared purpose.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
97% confidence
Finding

The curl-to-bash pipeline executes whatever script the endpoint returns without a separate verification step. HTTPS protects transport but does not prevent malicious changes at the distribution source; this is the same underlying installation risk identified in the other confirmed findings, not evidence of an additional independent exploit.

Content

Scanner excerpt · SKILL.md (reported line 56)May include surrounding context.

The skill exposes diagram generation despite describing its operations as searching and reading data.

Medium
Category
Excessive Agency
Confidence
88% confidence
Finding

The manifest at L03 describes searching and reading Excalidraw data, but create_view renders a diagram from supplied elements. Generating a new diagram is an output-creation capability rather than a search or retrieval operation, even if it does not persist changes remotely.

Content

No source excerpt is available for this finding.

Fallback setup executes downloaded installation scripts without warning about remote code execution or system changes.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

Both setup examples immediately execute code retrieved from a remote server without a pinned version, independent integrity verification, or explicit approval for local system changes. Restricting this behavior to a missing-CLI error reduces exposure, but a compromised distribution source could still execute arbitrary code with the user's permissions.

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 77)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