ClawHub
Back to skill

Security audit

Adr Writing

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward ADR-writing helper with a small local numbering script and no evidence of hidden, destructive, persistent, or network behavior.

Before installing, be aware that the skill may read nearby repository files and existing ADR filenames to provide context, then create or verify ADR markdown files under docs/adrs. Review generated ADRs for accuracy, especially when the source material is meeting notes or PR discussions.

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
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The description presents a document-authoring and validation skill focused on writing or formatting ADRs using MADR and checking ADR completeness against E.C.A.D.R. criteria. The actual code does none of that: it is a filesystem-oriented numbering helper for ADR filenames. Its primary purpose is materially different from the declared purpose, and it introduces undeclared capabilities involving directory discovery and file enumeration. While ADR numbering could support ADR creation workflows, this code chunk is not merely an implementation detail of formatting/quality checking; it is a separate utility with distinct triggers and behavior.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest explicitly narrows scope by saying this skill 'Does NOT extract decisions from conversations.' However, the documented usage includes 'Converting meeting notes or discussions to formal ADRs' and 'Recording technical choices from PR discussions,' which implies deriving decision content from conversational or discussion sources. That is a semantic mismatch between the declared boundary and the documented behavior of the skill.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The 'When To Use' section says the skill is for 'Converting meeting notes or discussions to formal ADRs' and 'Recording technical choices from PR discussions.' This conflicts with the manifest statement that the skill does not extract decisions from conversations. The contradiction is in the skill's own documentation about intended use, not merely an omitted detail.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.