ClawHub
Back to skill

Security audit

Fabric Bridge

Security checks for vulnerabilities and agentic risk

Overview

This is a legitimate Fabric CLI bridge, but it needs review because broad triggers could send user text, files, images, or URLs to external AI providers without an explicit privacy warning.

Install only if you are comfortable with Fabric AI using configured API keys and potentially sending provided text, files, images, URLs, or video-derived content to external AI providers. Use it deliberately for Fabric requests, redact sensitive data first, and prefer dry-run or approved local/private backends when handling confidential material.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (6)

External Model or Provider Selection

High
Category
Excessive Agency
Confidence
90% confidence
Finding

The skill explicitly documents external model/provider selection, confirming that user input can be routed to a third-party model endpoint. This is dangerous when combined with broad activation and missing disclosure because it increases the likelihood of sensitive content being sent off-box to external providers with different retention and policy controls.

Content

Scanner excerpt · SKILL.md (reported line 50)May include surrounding context.

Specify model

bash
echo "input" | fabric-ai -p <pattern> -m gpt-4o

Chinese output

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill description uses very broad activation language such as general requests for analysis, writing improvement, summarization, and mentions of a CLI. That can cause the skill to trigger for many ordinary user requests and silently route user content into an external tool, increasing the chance of unintended data disclosure or unexpected tool use.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation instructs users to run Fabric AI patterns on arbitrary input, but it does not clearly warn that content may be transmitted to external model providers via the fabric-ai CLI. Users may paste confidential text assuming local processing, which creates a real privacy and compliance risk.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The examples encourage processing YouTube URLs, web pages, images, and piped local file contents without any privacy, consent, or data-handling warning. In practice this makes it easy for users to send internal files, personal data, or third-party content to external services unintentionally, with greater exposure than plain text snippets.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
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.

Content

Scanner excerpt · SKILL.md (reported line 127)May include surrounding context.

md
## Custom Patterns

Create custom patterns at `~/.config/fabric/patterns/<name>/system.md`.

Each pattern directory contains a `system.md` file with the system prompt.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

The skill includes a dedicated 'Chinese output' example and guidance to use -g zh, which hard-codes a specific language option in the documentation. While not severe, this can be read as prescribing a locale behavior without explicitly framing it as user-selected output preference.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.