ClawHub
Back to skill

Security audit

Skill Doc Enhancer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local tool for analyzing and editing skill documentation, with notable care needed because it can overwrite a target SKILL.md file.

Install only if you want a local helper that can modify other skills' SKILL.md files. Prefer running with --dry-run first, review the generated content because it may contain placeholders, and apply changes only to skills you control under version control.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (7)

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The declared description says the skill both analyzes and enhances SKILL.md files, implying content generation or file modification. The supplied code performs only static inspection: it reads SKILL.md, inspects scripts/references/assets directories, checks for certain sections and placeholders, and outputs a JSON recommendation report. There is no code that writes to SKILL.md or any other file, no enrichment logic, and no restructuring/preservation workflow beyond reporting metrics. The analysis portion aligns with the description, but the primary enhancement capability is absent, making the description materially broader than the actual behavior.

Content

No source excerpt is available for this finding.

Self-Modification

High
Category
Rogue Agent
Confidence
94% confidence
Finding

The skill is explicitly designed to modify SKILL.md files, including guidance to edit another skill after running enhancement scripts. Self-modification or modification of other skills' instruction files is dangerous because it can alter agent behavior, inject persistence into future runs, and create a supply-chain style trust problem if changes are applied automatically without review.

Content

Scanner excerpt · SKILL.md (reported line 252)May include surrounding context.

python3 scripts/enhance_skill.py skills/nano-pdf

4. Review and edit manually if needed

Edit skills/nano-pdf/SKILL.md

text

## Notes

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
96% confidence
Finding

The skill instructs users to run local scripts that read from and modify files, but it does not declare any explicit tool scope such as allowed-tools or permissions. That creates an authorization and transparency gap: an agent may be induced to perform file reads and writes without clearly bounded access, increasing the risk of unintended modification of arbitrary skill directories or documentation files.

Content

No source excerpt is available for this finding.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% 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.

Content

Scanner excerpt · SKILL.md (reported line 252)May include surrounding context.

python3 scripts/enhance_skill.py skills/nano-pdf

4. Review and edit manually if needed

Edit skills/nano-pdf/SKILL.md

text

## Notes

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The top-level docstring and related function docstrings frame the behavior as automatic enhancement of documentation. In practice, lines L108-L120 and L157-L169 add placeholder text like 'Expected result here' and '[Use Case 1]', which contradicts the ordinary meaning of enhancement as improving substantive documentation.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest says the skill will analyze a skill directory and add usage examples, common use cases, and best practices, implying tailored enhancement based on the skill's contents. However, these functions generate static boilerplate such as 'Describe a common use case', '[Use Case 1]', and generic best-practice bullets, with only the scripts section doing any real analysis.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

This code performs a file write that overwrites the existing SKILL.md content, but the non-dry-run path provides no confirmation prompt and no nearby comment or docstring warning that the file will be changed in place. For a content-modifying operation, this lacks clear user disclosure at the point of action.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.