Back to skill

Security audit

Stiff-Sec — OpenClaw Security Hardener

Security checks for vulnerabilities and agentic risk

Overview

This skill is a mostly disclosed OpenClaw security hardener, but it can overwrite active security configuration using weak backup and restore handling while overstating some protections.

Review before installing. The skill is not malicious from the inspected artifacts, but it changes OpenClaw security settings and its restore path can overwrite your active config from an unvalidated backup. Use only in a controlled local environment after making your own independent backup, and do not rely on its claimed signed/hash lockfile protections without fixing the implementation.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (4)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/stiffen.py:56
Finding

Non-Atomic Replacement Can Corrupt the Active Security Configuration

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/stiffen.py:14
Finding

Predictable Backup Path Permits Backup Overwrite and Symlink Redirection

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/stiffen.py:75
Finding

Restore Operation Trusts Unvalidated and Potentially Symlinked Backup Entries

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
scripts/audit.py:19
Finding

Credential Scanner Discloses Secret Fragments in Console and Log Output

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

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 · .gitignore (reported line 4)May include surrounding context.

text
__pycache__/
*.pyc
*.pyo
.env
*.log
*.tmp
*.bak

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 · README.md (reported line 65)May include surrounding context.

text
⚠️  FOUND 1 potential plaintext credential(s):
   [channels.telegram.botToken] → value starts with: AAFNAA...ZOo
   Recommendation: replace with REDACTED or move to .env

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 · scripts/audit.py (reported line 88)May include surrounding context.

python
print(f"⚠️  FOUND {len(findings)} potential plaintext credential(s):")
        for path, preview in findings:
            print(f"   [{path}] → value starts with: {preview}")
        print("   Recommendation: replace with REDACTED or move to .env")
    else:
        print("✅ No plaintext credentials detected.")

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 · README.md (reported line 98)May include surrounding context.

md
## Who Is This For?

- OpenClaw operators running local or LAN-exposed gateways
- Anyone who's ever had a skill modify their config without asking
- Anyone whose ADHD moves faster than their safety checks 😅
- Anyone who wants to know *exactly* what state their install is in

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 9)May include surrounding context.

md
Generic auditors just "tell" you there's a problem. **Stiff-Sec** (the Hex-Stiff) slams the door, locks the vault, and hardens your OpenClaw setup against prying eyes.

## Our Backup Policy (Sienna's Protocol)
- **Always Backup First:** Before any config change, we create a timestamped backup in `~/.openclaw/backups/`.
- **Undo Strategy:** Any change can be reversed with `stiff-sec restore`.
- **Explain Everything:** Every edit is logged in `MEMORY.md` with a specific "Undo" instruction.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The restore path blindly copies the most recent backup over the active user configuration without any confirmation, validation, or integrity check. In an agent/tooling context, this can unexpectedly revert security settings or restore a stale or attacker-planted configuration if the backup directory contents are untrusted or manipulated.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The section header 'Lock File' suggests file-locking behavior to prevent concurrent access, but the code simply writes a .stiffened marker file. This is an intent-level documentation mismatch because the comment describes a synchronization/locking action that is not what the implementation performs.

Content

No source excerpt is available for this finding.