Back to skill

Security audit

Telegraph Protocol

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly upfront about being a paid Telegraph inference integration, but it grants automatic wallet-spending authority and a mutable remote tool surface with weak setup controls.

Install only with a strictly limited burner wallet, assume paid calls may spend USDC automatically, and avoid sending sensitive prompts or media unless you trust the Telegraph MCP package, its live registry, and the configured endpoints. Prefer a pinned package version, HTTPS endpoints, explicit spending limits, and confirmation before any paid call.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Tool Hijacking and SpoofingModifies or replaces tools so legitimate-looking calls execute attacker logic
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Findings (4)

T08 · Insecure Dependencies

Error
Location
SKILL.md:139
Finding

Unpinned npm Package Executes with Access to Wallet Credentials

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:121
Finding

Payment and Inference Operations Use Unencrypted Bare-IP HTTP Endpoints

Content
View full analysis
Remediation
View remediation

T07 · Tool Hijacking and Spoofing

Error
Location
SKILL.md:85
Finding

Remote Registry Can Change the Available Tool Surface Without Local Approval

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
SKILL.md:121
Finding

Paid Calls Sign Wallet Authorizations Without Agent-Visible Approval

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The activation text is broad enough to trigger on very common user requests like weather, climate data, LLM completions, embeddings, image generation, or signal monitoring, even when the user did not specifically ask to use Telegraph. In this skill, activation can lead to paid tool use and external routing through a decentralized network, so over-broad triggering creates a real risk of unnecessary third-party disclosure and unintended spending.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill states that the MCP server handles x402 payments transparently and that the agent never sees the payment step, but the user-facing description does not make clear that ordinary inference requests may automatically sign and spend USDC from the configured wallet. Because many listed use cases are common assistant tasks, this can cause silent or surprising monetary loss if the skill activates without clear user consent.

Content

No source excerpt is available for this finding.