Back to skill

Security audit

Defuddle Obsidian

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward webpage-to-markdown helper, with the main caution that it asks users to install an unpinned third-party CLI globally.

Install only if you are comfortable using the Defuddle npm package to fetch and convert web pages. Prefer a reviewed, pinned, project-local install when possible, and do not run the npm install command with elevated privileges.

Vulnerability Patterns
  • 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
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:10
Finding
Unpinned Global Installation of a Third-Party npm Package## Vulnerability Details **File Location**: `SKILL.md`, lines 10–12 **Vulnerability Type**: Unpinned third-party dependency installed globally **Risk Level**: Medium **Vulnerable code snippet**: ```markdown If not installed: `npm install -g defuddle` ``` ### Technical Analysis The Skill instructs users or agents to install the latest available version of `defuddle` from the configured npm registry. The command does not pin a reviewed version or verify package integrity. Consequently, the package contents executed during installation can differ from those assessed when the Skill was reviewed. npm installation may execute package lifecycle scripts such as `preinstall`, `install`, or `postinstall`. If the package, one of its dependencies, the registry account, or registry resolution is compromised, those scripts can execute arbitrary commands with the privileges of the user running npm. The `-g` option also installs the package globally. This exceeds the minimum scope needed to process web pages because a project-local, version-pinned installation would be sufficient. Global installation can modify user-level global npm directories and expose executables across unrelated projects. The instruction does not explicitly request elevated privileges, so root-level impact depends on whether the operator independently runs it with administrative permissions. The separate instructions to run `defuddle parse <url>` necessarily cause network access to retrieve a user-selected webpage. That behavior is consistent with the declared webpage-extraction functionality, and the reviewed file does not instruct Defuddle to collect credentials or upload local sensitive files. Therefore, the pre-scan network alert is not confirmed as data exfiltration. ### Attack Path 1. An attacker compromises the `defuddle` npm publisher account, a transitive dependency, or the registry used by the environment. 2. The attacker publishes or serves a maliciou ...[truncated 1198 chars]
Remediation
## Remediation Suggestions - Pin Defuddle to a specifically reviewed version instead of resolving the mutable latest release. - Prefer a project-local dependency recorded in `package.json` and a committed lockfile rather than using `-g`. - Use deterministic installation, such as `npm ci`, with lockfile integrity metadata. - Review the selected package version, its transitive dependencies, and any lifecycle scripts before permitting installation. - Where lifecycle scripts are unnecessary, consider installing with `--ignore-scripts`, subject to compatibility testing. - Configure npm to use an approved registry and validate package provenance or signatures where supported. - Run the parser under a non-privileged account or sandbox with access limited to the required destination URL and output location. - Document that operators must not use `sudo` or another privilege-elevation mechanism for installation.
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 (1)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The description says to use this skill whenever the user provides a URL to read or analyze, including broad categories like articles, blog posts, or any standard web page. That scope is ambiguous and lacks constraints or negative examples, which could cause unintended invocation for routine browsing-related requests.

Static analysis

No suspicious patterns detected.