Back to skill

Security audit

Full Stack Feature

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed feature-development workflow guide, but its install instructions use mutable remote sources that users should pin before running.

Review and pin the installer before use: prefer a fixed clawhub version or immutable Git commit instead of @latest or a mutable branch. Once installed, expect this skill to guide an agent through broad codebase changes for feature work, so use it in a repository where you are comfortable allowing user-directed implementation, testing, and documentation edits.

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:16
Finding
Unpinned Remote Dependencies Executed Through npx## Vulnerability Details **File Location**: `SKILL.md:16-18`, `README.md:24-32` **Vulnerability Type**: T08: Insecure Dependencies **Risk Level**: Medium ### Vulnerable Code `SKILL.md:16-18`: ```bash npx clawhub@latest install full-stack-feature ``` `README.md:24-32`: ```bash npx add https://github.com/wpank/ai/tree/main/skills/meta/full-stack-feature ``` ```bash npx clawhub@latest install full-stack-feature ``` ### Technical Analysis The documented installation procedures use `npx` to download and execute remote npm packages without immutable version or integrity constraints. In particular, `clawhub@latest` explicitly selects a mutable package release. The `npx add` command also invokes a remotely resolved npm package while passing a GitHub source located on a mutable branch. No exact package version, Git commit hash, package-lock constraint, integrity checksum, or cryptographic signature is provided. Consequently, the code executed by these commands can change after this skill has been reviewed. If the npm package, publisher account, release process, or referenced GitHub repository is compromised, an attacker could distribute modified installer code to subsequent users. This is a supply-chain risk rather than evidence that the currently referenced packages are malicious. ### Attack Path 1. An attacker compromises the npm publisher account, package release pipeline, or GitHub repository referenced by the documentation. 2. The attacker publishes a malicious package version selected by `@latest`, modifies the mutable repository branch, or otherwise alters a remotely resolved installer dependency. 3. A user follows one of the documented installation commands. 4. `npx` downloads the changed package and executes its command or applicable lifecycle behavior with the user's privileges. 5. The malicious installer can perform actions available to that user, such as modifying project or user files, reading acc ...[truncated 748 chars]
Remediation
## Remediation Suggestions 1. Replace `clawhub@latest` with an explicitly reviewed, immutable package version. 2. Pin repository-based installation to a specific Git commit rather than a mutable branch such as `main`. 3. Publish and verify integrity checksums or cryptographic signatures for distributed artifacts. 4. Document the expected package publisher, source repository, version, and checksum so users can verify provenance before installation. 5. Prefer a reviewed local or manual installation method when feasible. 6. In CI environments, use lockfiles, approved registries, dependency allowlists, and restricted credentials. 7. Run installation in a sandbox or least-privileged environment without unrelated secrets. 8. Review installer and lifecycle scripts for every dependency update before changing the pinned version.
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Rp1

Medium
Category
MCP Rug Pull
Confidence
91% confidence
Finding
The README instructs users to run `npx add` against a GitHub URL without pinning to an immutable version, tag, or commit. That creates a supply-chain risk because future changes to the referenced repository can alter what gets installed, so users may execute unreviewed or malicious content later even if the current repository state is benign.

Rp1

Medium
Category
MCP Rug Pull
Confidence
89% confidence
Finding
Using `npx clawhub@latest install full-stack-feature` pulls and executes the latest package version at runtime, which is mutable and may change without notice. If the package or its dependency chain is compromised, users can be exposed to arbitrary code execution during installation.

Session Persistence

Medium
Category
Rogue Agent
Content
From your project root:

```bash
mkdir -p .cursor/skills
cp -r ~/.ai-skills/skills/meta/full-stack-feature .cursor/skills/full-stack-feature
```
Confidence
60% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Skill Enumeration

Medium
Category
Agent Snooping
Content
From your project root:

```bash
mkdir -p .claude/skills
cp -r ~/.ai-skills/skills/meta/full-stack-feature .claude/skills/full-stack-feature
```
Confidence
85% confidence
Finding
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Skill Enumeration

Medium
Category
Agent Snooping
Content
#### Claude Code (global)

```bash
mkdir -p ~/.claude/skills
cp -r ~/.ai-skills/skills/meta/full-stack-feature ~/.claude/skills/full-stack-feature
```
Confidence
85% confidence
Finding
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

External Model or Provider Selection

Medium
Category
Excessive Agency
Content
---
name: full-stack-feature
model: reasoning
description: Meta-skill for orchestrating end-to-end feature development from persona research through deployed, tested code — coordinating skills, commands, and agents across every stage. Use when building a new feature end-to-end, coordinating frontend + backend work, or needing a structured approach from research to production.
---
Confidence
90% confidence
Finding
Skill selects an external model or provider that may use a different account or billing plan than the operator expects. Undisclosed model switches can cause unexpected cost or quota consumption.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
The installation command uses `npx clawhub@latest`, which pulls whatever version is current at execution time rather than a reviewed, fixed release. That creates a supply-chain risk: a compromised or breaking upstream release could be executed by users immediately, and because this is an install path in a meta-skill, it is likely to be copied and run as-is.

Static analysis

No suspicious patterns detected.