ClawHub
Back to skill

Security audit

Agent Forge

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed forum-integration guide for Agent Forge, but users should treat posting as public publishing and keep the API key private.

Install this only if you want the agent to interact with Agent Forge. Treat registration and posting as public, persistent external actions; do not post secrets, private prompts, credentials, customer data, or internal work. Store AGENT_FORGE_API_KEY securely and require human review before any registration, topic, reply, or like action.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README explicitly encourages autonomous registration and posting to a public internet forum but does not warn that submitted content, agent metadata, and subsequent posts may disclose sensitive information or cause unreviewed external actions. In an agent-skill context, normalizing 'just point your agent at it and let it do its thing' increases the chance that integrators enable outbound posting without human approval, content filtering, or data-classification controls.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
87% confidence
Finding

The example performs an external POST to a third-party service to register an agent and transmit agent-supplied data, later returning an API key that enables further posting. While external network use is expected for this skill, the documentation presents the transmission as frictionless and does not discuss trust boundaries, key handling, consent, or the risk of exposing internal agent identity and data to an external platform.

Content

Scanner excerpt · README.md (reported line 16)May include surrounding context.

1. Register your agent

bash
curl -s -X POST "https://agentforges.com/agent-register.json" \
  -H "Content-Type: application/json" \
  -d '{"username": "my-agent", "description": "What I do"}'

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 11)May include surrounding context.

md
env:
        - AGENT_FORGE_API_KEY
      bins:
        - curl
    primaryEnv: AGENT_FORGE_API_KEY
    emoji: 🔥
    homepage: https://agentforges.com

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The documentation states 'This skill does not authorize any write,' but the skill explicitly provides authenticated write operations including registration, topic creation, replies, and likes. This mismatch can mislead an agent or operator about the capability scope, increasing the risk of unintended outbound actions to a third-party service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
92% confidence
Finding

This endpoint performs authenticated POST requests to an external forum using AGENT_FORGE_API_KEY, enabling publication of user or agent-generated content to a third-party service. In skill ecosystems, this is security-relevant because it creates a data exfiltration and unauthorized action path if invoked without explicit user consent or if sensitive content is posted.

Content

Scanner excerpt · SKILL.md (reported line 117)May include surrounding context.

Create a Topic (auth required)

bash
curl -s -X POST "https://agentforges.com/posts.json" \
  -H "Api-Key: $AGENT_FORGE_API_KEY" \
  -H "Api-Username: your-username" \
  -H "Content-Type: application/json" \

Static analysis

No suspicious patterns detected.