ClawHub
Back to skill

Security audit

harness

Security checks for vulnerabilities and agentic risk

Overview

This is a plain-text workflow skill for planning, guardrails, and recovery, with no hidden code, persistence, or data exfiltration found.

Installers should expect this skill to make agents follow a more deliberate workflow, ask clarifying questions, read relevant code or docs, and block destructive commands unless explicitly authorized. Consider narrowing activation phrases if accidental invocation would be disruptive, and review the declared code-workflow dependency separately.

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
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • 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
Findings (4)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
85% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · guardrails.md (reported line 25)May include surrounding context.

md
| Class | Examples | Default behavior |
|-------|----------|------------------|
| Filesystem | `rm -rf`, `rm -f /`, `find … -delete` (broad) | Block — require explicit user authorization with the exact path |
| Git | `git push --force`, `git reset --hard`, `git checkout -- .`, `git clean -f`, `git branch -D` | Block — require explicit user authorization |
| Database / state | `DROP TABLE`, `TRUNCATE`, `rm -rf <data-dir>` | Block — require explicit user authorization |
| Process | `kill -9 <pid>` against unknown processes | Block — require explicit user authorization |

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
70% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · guardrails.md (reported line 26)May include surrounding context.

md
| Class | Examples | Default behavior |
|-------|----------|------------------|
| Filesystem | `rm -rf`, `rm -f /`, `find … -delete` (broad) | Block — require explicit user authorization with the exact path |
| Git | `git push --force`, `git reset --hard`, `git checkout -- .`, `git clean -f`, `git branch -D` | Block — require explicit user authorization |
| Database / state | `DROP TABLE`, `TRUNCATE`, `rm -rf <data-dir>` | Block — require explicit user authorization |
| Process | `kill -9 <pid>` against unknown processes | Block — require explicit user authorization |
| Hooks bypass | `--no-verify`, `--no-gpg-sign`, `--force` against `main`/`master` | Block — require explicit user authorization |

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
65% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · guardrails.md (reported line 26)May include surrounding context.

md
| Class | Examples | Default behavior |
|-------|----------|------------------|
| Filesystem | `rm -rf`, `rm -f /`, `find … -delete` (broad) | Block — require explicit user authorization with the exact path |
| Git | `git push --force`, `git reset --hard`, `git checkout -- .`, `git clean -f`, `git branch -D` | Block — require explicit user authorization |
| Database / state | `DROP TABLE`, `TRUNCATE`, `rm -rf <data-dir>` | Block — require explicit user authorization |
| Process | `kill -9 <pid>` against unknown processes | Block — require explicit user authorization |
| Hooks bypass | `--no-verify`, `--no-gpg-sign`, `--force` against `main`/`master` | Block — require explicit user authorization |

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This manifest description says to use the skill when 'enforcing stable AI agent workflows' or 'recovering from verification failures' and lists short trigger phrases like 'harness' and 'AI harness'. Those terms are broad enough to overlap with ordinary discussion of AI workflows, and the file does not provide exclusion conditions or negative examples to clarify when the skill should not activate.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.