Back to skill

Security audit

CommPeak

Security checks for vulnerabilities and agentic risk

Overview

This skill needs review because it treats messaging credentials as ordinary reads and its setup can make broader changes than it explains.

Review CLI installation separately and prefer an independently verified release. Consider disabling automatic skill synchronization with OO_SKILLS_SYNC_DISABLED=1. Keep returned stream tokens out of responses and logs; prefer send_sms, which retrieves its token internally, and confirm recipients and message content. The remote installer and hosted connector were not verified, and no artifact-backed 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 (5)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied command performs remote installer execution, not CommPeak data operations. Nothing in the supplied code establishes that installing this CLI supports the declared CommPeak functionality. The remote script's contents are not provided, so its additional effects cannot be assessed.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

The command streams a remotely supplied script directly into Bash without a pinned version or independent integrity verification; the PowerShell alternative uses the same execution pattern. If the distribution endpoint or release infrastructure is compromised, attacker-controlled code executes with the agent process's privileges, potentially exposing local data and credentials; HTTPS alone does not prevent publisher-side compromise.

Content

Scanner excerpt · SKILL.md (reported line 62)May include surrounding context.

The claim that agents never handle raw tokens contradicts the exposed stream-token retrieval action.

Medium
Category
Excessive Agency
Confidence
92% confidence
Finding

The skill promises that agents never handle raw tokens, yet exposes get_stream_token and classifies untagged reads as safe to run directly. Although retrieving a token can be legitimate, this guidance lacks secret-handling safeguards and can unnecessarily expose a messaging credential in agent context, logs, or responses; no explicit exfiltration destination is present.

Content

No source excerpt is available for this finding.

The setup workflow executes downloaded installation scripts without warning the user about remote code execution and potential system changes.

Medium
Category
Output Handling
Confidence
93% confidence
Finding

When the CLI is missing, the setup instructions direct the agent to install executable software without a separate approval step or explanation of local system changes. This expands a CommPeak data task into host modification; the missing-CLI condition limits exposure but does not establish user authorization for installation.

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 83)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