ClawHub
Back to skill

Security audit

Venture Spawner

Security checks for vulnerabilities and agentic risk

Overview

This skill is an agent-spawning workflow helper, but it delegates git pushes, SSH-related outreach work, and sensitive context without clear approval or least-privilege controls.

Review before installing. This skill is not deceptive or destructive by itself, but it is meant to coordinate autonomous sub-agents that may modify repositories, push commits, use SSH-related instructions, and handle sensitive business or personal files. Use it only in a controlled workspace with trusted JOB_BOARD.md authors, limited credentials, protected branches, and explicit approval before any push, remote command, or sensitive-data handling.

A.I.G

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:61
Finding

Untrusted Job Content Is Copied into Privileged Sub-Agent Instructions

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Error
Location
SKILL.md:42
Finding

Sensitive Context and High-Impact Git or SSH Operations Are Delegated without Least-Privilege Controls

Content
View full analysis
Remediation
View remediation

SkillSpector

Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 77)May include surrounding context.

Constraints

  • [Bucket-specific constraints from table above]
  • Do NOT modify .env files with real keys
  • Build must pass before committing
  • Git commit with descriptive message
text

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly instructs spawned sub-agents to perform git commits and, for BUILD jobs, to push changes, but it provides no user-facing warning or consent checkpoint before initiating those actions. In an agent-spawning skill, this increases the risk of unintended repository modification or publication because downstream agents may act autonomously based on orchestrator input rather than explicit user approval.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 162)May include surrounding context.

md
- ❌ Spawning 10 agents at once (max 3)
- ❌ Spawning an agent for a 2-minute task (do it inline)
- ❌ Not passing context (agent wastes time re-reading)
- ❌ Marking ✅ without checking output exists
- ❌ Polling subagents in a loop (wait for push events)
- ❌ Spawning duplicate agents for the same job
- ❌ Not killing stale agents (> timeout + 5 min buffer)