openclaw/docs/compliance/vulnerability-disclosure.md

---
title: Vulnerability Disclosure Policy
summary: Responsible disclosure policy for reporting security vulnerabilities in Clawdbot.
permalink: /compliance/vulnerability-disclosure/
---

# Vulnerability Disclosure Policy

This policy describes how to report security vulnerabilities in Clawdbot and how we handle disclosures.

## Reporting a Vulnerability

### Preferred Channels

**GitHub Security Advisories (Recommended)**

Report vulnerabilities privately through GitHub Security Advisories:

1. Go to [Clawdbot Security Advisories](https://github.com/clawdbot/clawdbot/security/advisories)
2. Click "Report a vulnerability"
3. Provide details following the template below

**Email**

For reporters who cannot use GitHub:
- Email: security@clawd.bot
- Encrypt sensitive details using our PGP key (available at the repository)

### What to Include

Your report should include:

| Field | Description |
|-------|-------------|
| **Summary** | Brief description of the vulnerability |
| **Severity** | Your assessment: Critical, High, Medium, Low |
| **Affected versions** | Which Clawdbot versions are affected |
| **Attack vector** | How the vulnerability can be exploited |
| **Impact** | What an attacker could achieve |
| **Reproduction steps** | Step-by-step instructions to reproduce |
| **Proof of concept** | Code, screenshots, or logs demonstrating the issue |
| **Suggested fix** | If you have recommendations |

### Example Report

```
## Summary
Authentication bypass in gateway pairing mechanism

## Severity
High

## Affected Versions
Clawdbot 2024.1.0 through 2024.1.15

## Attack Vector
Remote, requires network access to gateway

## Impact
Attacker could pair a device without valid pairing code under specific timing conditions

## Reproduction Steps
1. Start Clawdbot gateway with default configuration
2. Initiate pairing request from attacker device
3. Send rapid concurrent pairing attempts with invalid codes
4. Under race condition, one attempt may succeed

## Proof of Concept
[Attached script or detailed steps]

## Suggested Fix
Add mutex lock around pairing code validation
```

## Response Timeline

| Phase | Timeline | Description |
|-------|----------|-------------|
| Acknowledgment | 2 business days | Confirm receipt of report |
| Initial assessment | 5 business days | Determine validity and severity |
| Status update | 10 business days | Provide fix timeline or request more info |
| Fix development | Varies by severity | Develop and test remediation |
| Coordinated disclosure | 90 days max | Public disclosure after fix available |

### Severity-Based Response

| Severity | Target Fix Time | Disclosure Timeline |
|----------|-----------------|---------------------|
| Critical | 7 days | 14 days after fix |
| High | 30 days | 30 days after fix |
| Medium | 60 days | 60 days after fix |
| Low | 90 days | 90 days after fix |

## Safe Harbor

We want security researchers to feel comfortable reporting vulnerabilities. If you:

- Act in good faith to avoid privacy violations, data destruction, and service disruption
- Only interact with accounts you own or have explicit permission to test
- Do not exploit vulnerabilities beyond what is necessary to demonstrate the issue
- Report vulnerabilities promptly and do not disclose publicly before coordinated disclosure
- Do not demand payment for vulnerability information

Then we commit to:

- Not pursuing legal action against you for security research conducted in accordance with this policy
- Working with you to understand and resolve the issue quickly
- Recognizing your contribution (with your permission) in our security acknowledgments
- Keeping you informed of our progress toward remediation

## Scope

### In Scope

| Component | Description |
|-----------|-------------|
| Clawdbot CLI | Core command-line application |
| Gateway | WebSocket server and HTTP endpoints |
| Channel integrations | WhatsApp, Telegram, Discord, Slack, Signal, iMessage |
| Plugins | Official Clawdbot plugins |
| Configuration | Security of config files and credentials |
| Authentication | Gateway auth, pairing, RBAC |

### Out of Scope

| Component | Reason |
|-----------|--------|
| Third-party services | Report to respective providers (OpenAI, Anthropic, etc.) |
| Messaging platform bugs | Report to WhatsApp, Telegram, Discord, etc. |
| User configuration errors | Covered by security audit and hardening guide |
| Social engineering | Not a software vulnerability |
| Physical attacks | Out of scope for software security |
| Denial of service | Unless it reveals an amplification vector |

### Qualifying Vulnerabilities

| Category | Examples |
|----------|----------|
| Authentication bypass | Accessing gateway without valid credentials |
| Authorization bypass | Accessing resources beyond granted permissions |
| Injection attacks | Command injection, prompt injection with security impact |
| Information disclosure | Credential leakage, sensitive data exposure |
| Cryptographic issues | Weak encryption, improper key handling |
| Logic flaws | Security control bypass, race conditions |

### Non-Qualifying Issues

| Category | Examples |
|----------|----------|
| Best practices | Missing headers without demonstrated impact |
| Self-XSS | Requires victim to inject malicious content |
| Rate limiting bypass | Without security impact |
| Version disclosure | Unless it reveals specific vulnerable version |
| Already reported | Duplicate of known issue |

## Coordinated Disclosure

We follow coordinated disclosure principles:

1. **Reporter contacts us** with vulnerability details
2. **We acknowledge** receipt within 2 business days
3. **We assess** severity and develop fix
4. **We coordinate** disclosure date with reporter
5. **We release** fix and publish advisory
6. **Reporter may publish** after coordinated disclosure date

### Disclosure Content

Our security advisories include:
- Vulnerability description
- Affected versions
- Fixed versions
- Severity rating (CVSS if applicable)
- Mitigation steps
- Credit to reporter (if desired)

### Early Disclosure Exceptions

We may disclose earlier if:
- Vulnerability is being actively exploited
- Public disclosure is imminent from another source
- Significant user harm can be prevented

We may request delayed disclosure if:
- Fix requires significant development time
- Coordinating with dependent projects
- Critical infrastructure implications

## Recognition

We maintain a Security Hall of Fame for researchers who report valid vulnerabilities:

- Acknowledgment in security advisory
- Listed in SECURITY.md (with permission)
- Letter of appreciation (on request)

We do not currently offer monetary bounties, but we deeply appreciate responsible disclosure.

## Legal

This policy is not a license to conduct security research on systems you do not own or have permission to test.

Always obtain proper authorization before testing. This policy applies only to:
- Your own Clawdbot installations
- Test environments you control
- Systems where you have explicit written permission

Unauthorized access to computer systems is illegal. This policy does not authorize any activity that would violate applicable laws.

## Contact

- **Security reports**: [GitHub Security Advisories](https://github.com/clawdbot/clawdbot/security/advisories) (preferred)
- **Email**: security@clawd.bot
- **General questions**: Use GitHub Issues (do not disclose vulnerabilities publicly)

## Policy Updates

This policy may be updated periodically. Significant changes will be announced in the repository.

## Related Documentation

- [Threat Model](/security/threat-model) - Security architecture and known risks
- [Incident Response](/compliance/incident-response) - How we handle security incidents
- [Security Hardening](/enterprise/security-hardening) - Secure configuration guide

---

*Policy owner: Security Team*
*Effective date: 2026-01-27*
*Last reviewed: 2026-01-27*