Back to skill

Security audit

Competitor Teardown

Security checks for vulnerabilities and agentic risk

Overview

The skill’s competitive-research purpose is coherent, but its setup instructions ask users to execute mutable remote installers and unpinned skill installs.

Review this skill before installing. Use a pinned, independently verified installer for the inference.sh CLI, avoid running `curl | sh`, run login as a separate step, and do not install the optional related skills unless you have reviewed and pinned their exact versions.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (2)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:14
Finding

Unpinned Remote Installer Executed Directly by Shell

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:306
Finding

Unpinned Third-Party Skill Installation Through npx

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • 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
Findings (4)

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

curl -fsSL https://cli.inference.sh | sh pipes network-fetched content directly into the shell, which is a classic arbitrary code execution pattern. Even with the nearby note about checksum verification, the shown one-liner executes before any independent user verification, so compromise of the hosting endpoint, DNS/TLS path, or served script could lead to immediate code execution on the user's machine.

Content

Scanner excerpt · SKILL.md (reported line 14)May include surrounding context.

Quick Start

bash
curl -fsSL https://cli.inference.sh | sh && infsh login

# Research competitor landscape
infsh app run tavily/search-assistant --input '{

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest description includes a long list of broad trigger phrases such as market research, competitor comparison, and review analysis, which can cause the skill to activate for many generic business requests. Over-broad activation increases the chance that users are steered into using this skill's riskier instructions, including external installs and browser/search tooling, in contexts where it was not specifically intended.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The skill recommends invoking npx skills add ... without pinning an exact package/version, which allows whatever package version is current at execution time to run code during installation. This creates a supply-chain risk: if the package is updated maliciously, compromised, or publishes a breaking change, users may execute unreviewed code simply by following the skill instructions.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

This second npx skills add ... command has the same issue: it executes a package resolution path that is not pinned to a specific trusted version. In a skill file, these copy-pasteable setup commands are especially risky because users may run them directly without reviewing what code will be fetched and executed.

Content

No source excerpt is available for this finding.