Back to skill

Security audit

Autonomous Execution

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only autonomy guidance skill that tells the agent to complete in-scope tasks while stopping for sensitive actions.

This skill is suitable for users who want the agent to keep working through ordinary in-scope tasks, but you should still review prompts carefully because it encourages more autonomous follow-through before asking for help.

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
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Credential Access

High
Category
Privilege Escalation
Content
### Commitment Contract
- Complete all subtasks within scope
- Ask for confirmation on sensitive operations
- Never access secrets or credentials
- Report partial results if must stop

### Safe Completion Checklist
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Credential Access

High
Category
Privilege Escalation
Content
| Multiple items to process | Complete all within scope |
| Error on read operation | Retry, then ask |
| Error on auth/credentials | STOP - ask user |
| Need to access secrets | STOP - ask user |
| Need to send message | STOP - ask user |
| Task requires sensitive action | STOP - ask user |
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Content
- Don't read outside workspace without permission
- Don't try to bypass auth errors
- Don't access environment variables or secrets
- Don't modify system files
- Don't make changes outside the task scope

### Error Flow
Confidence
60% confidence
Finding
Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Static analysis

No suspicious patterns detected.