Security Audit

mike-coulbourn/claude-vibes:plugins/vibes/skills/skill-builder

github.com/mike-coulbourn/claude-vibes

AI SkillCommit b6e9c9a1a15f

CRITICAL

Scanned 3 days ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

mike-coulbourn/claude-vibes:plugins/vibes/skills/skill-builder received a trust score of 0/100, placing it in the Untrusted category. This skill has significant security findings that require attention before use in production.

SkillShield's automated analysis identified 10 findings: 4 critical, 5 high, 1 medium, and 0 low severity. Key findings include File read + network send exfiltration, Sensitive path access: AI agent config, Command Injection via unsanitized placeholders in shell commands.

The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. The Manifest Analysis layer scored lowest at 0/100, indicating areas for improvement.

Last analyzed on March 22, 2026 (commit b6e9c9a1). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

Static Code Analysis

40%

Dependency Graph

100%

LLM Behavioral Safety

78%

Behavioral Risk Signals

Network Access

4 findings

Filesystem Write

8 findings

Shell Execution

6 findings

Dynamic Code

1 finding

Excessive Permissions

8 findings

Security Findings10

Severity	Finding	Layer	Location
CRITICAL	File read + network send exfiltration AI agent config/credential file access Remove access to sensitive files not required by the skill's stated purpose. SSH keys, cloud credentials, and browser data should never be read by skills unless explicitly part of their declared functionality.	Manifest	plugins/vibes/skills/skill-builder/SKILL.md:30
CRITICAL	File read + network send exfiltration AI agent config/credential file access Remove access to sensitive files not required by the skill's stated purpose. SSH keys, cloud credentials, and browser data should never be read by skills unless explicitly part of their declared functionality.	Manifest	plugins/vibes/skills/skill-builder/SKILL.md:107
CRITICAL	File read + network send exfiltration AI agent config/credential file access Remove access to sensitive files not required by the skill's stated purpose. SSH keys, cloud credentials, and browser data should never be read by skills unless explicitly part of their declared functionality.	Manifest	plugins/vibes/skills/skill-builder/SKILL.md:113
CRITICAL	File read + network send exfiltration AI agent config/credential file access Remove access to sensitive files not required by the skill's stated purpose. SSH keys, cloud credentials, and browser data should never be read by skills unless explicitly part of their declared functionality.	Manifest	plugins/vibes/skills/skill-builder/SKILL.md:213
HIGH	Sensitive path access: AI agent config Access to AI agent config path detected: '~/.claude/'. This may indicate credential theft. Verify that access to this sensitive path is justified and declared.	Static	plugins/vibes/skills/skill-builder/SKILL.md:30
HIGH	Sensitive path access: AI agent config Access to AI agent config path detected: '~/.claude/'. This may indicate credential theft. Verify that access to this sensitive path is justified and declared.	Static	plugins/vibes/skills/skill-builder/SKILL.md:107
HIGH	Sensitive path access: AI agent config Access to AI agent config path detected: '~/.claude/'. This may indicate credential theft. Verify that access to this sensitive path is justified and declared.	Static	plugins/vibes/skills/skill-builder/SKILL.md:113
HIGH	Sensitive path access: AI agent config Access to AI agent config path detected: '~/.claude/'. This may indicate credential theft. Verify that access to this sensitive path is justified and declared.	Static	plugins/vibes/skills/skill-builder/SKILL.md:213
HIGH	Command Injection via unsanitized placeholders in shell commands The skill provides instructions for the agent (Claude) to execute shell commands that include placeholders like `skill-name` and `required-package`. If the agent substitutes untrusted user input directly into these placeholders without proper sanitization, it could lead to command injection. For example, if a user provides `my-skill; rm -rf /` as a skill name, the `mkdir` command could become `mkdir -p ~/.claude/skills/my-skill; rm -rf /`. This pattern is present in multiple commands throughout the skill, including `mkdir`, `python scripts/helper.py`, `pip install`, and `cat` commands. Instruct the agent to sanitize all user-provided inputs before incorporating them into shell commands. Specifically, escape or quote arguments, and validate input against expected patterns (e.g., only alphanumeric and hyphens for skill names). Ensure that the agent's execution environment for these commands is sandboxed or has minimal necessary permissions.	LLM	SKILL.md:100
MEDIUM	Unpinned dependency installation The skill instructs the agent to use `pip install required-package` for installing dependencies. This command does not specify a version, leading to unpinned dependencies. This can introduce supply chain risks, as a future version of `required-package` could be malicious or contain vulnerabilities, or a typosquatted package could be installed if the name is misspelled or user input is not validated. Always specify exact versions for dependencies (e.g., `pip install required-package==1.2.3`) or use a lock file mechanism (e.g., `pip install -r requirements.txt` where `requirements.txt` has pinned versions). Instruct the agent to validate package names and versions against a trusted source or manifest.	LLM	SKILL.md:169

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/da74610138cd3e5a.svg)](https://skillshield.io/report/da74610138cd3e5a)