Back to skill

Security audit

url-to-markdown

Security checks for vulnerabilities and agentic risk

Overview

This skill is a read-only helper for converting public web pages to Markdown, with clear limits and no evidence of hidden privileged behavior.

Install only if you are comfortable sending public page URLs through ReplyNodes' Markdown endpoint. Do not use it for private, local, login-only, credential-bearing, or sensitive URLs; the skill itself appropriately tells the agent to refuse those cases.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (3)

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 53)May include surrounding context.

md
`https://example.com/path?foo=bar`.

No API key, account, credential, MCP server, paid plan, or hidden telemetry is
required. Never ask the user for credentials for this workflow. Keep the exact
input URL alongside the returned Markdown; do not rewrite, shorten,
canonicalize, or replace the source URL.

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 87)May include surrounding context.

md
`https://example.com/path?foo=bar`.

No API key, account, credential, MCP server, paid plan, or hidden telemetry is
required. Never ask the user for credentials for this workflow. Keep the exact
input URL alongside the returned Markdown; do not rewrite, shorten,
canonicalize, or replace the source URL.

Unbounded Resource Access

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill allows unbounded resource consumption (API calls, storage, compute). Without rate limits or quotas, a compromised or misbehaving agent can cause denial-of-service or cost overruns.

Content

Scanner excerpt · SKILL.md (reported line 97)May include surrounding context.

md
websites, publish content, schedule work, or call social/provider write APIs.
- If the URL is unsupported, malformed, unreachable, blocked, or exceeds the
  service's limits, report a concise bounded failure with the original URL and
  the service error/status when available. Do not retry indefinitely, bypass a
  site restriction, or fall back to private access.
- If extraction is partial, label it partial and return only the content
  received; never invent missing text or claim that JavaScript/private content