Contributing to the OpenClaw Threat Model

Learn how to add threats or mitigations to the OpenClaw threat model, even without security experience. This guide covers the contribution process and what details to include.

Read this when

  • You want to contribute security findings or threat scenarios
  • Reviewing or updating the threat model

The threat model evolves over time. Anyone can contribute, and prior experience with security or MITRE ATLAS is not a prerequisite.

Note

This process covers additions to the threat model, not live vulnerability reports. If you have found an exploitable flaw, use the responsible-disclosure process described on the Trust page instead.

Ways to contribute

Introduce a threat. File a documentation issue and describe the attack scenario in your own terms. The following details are useful but optional:

  • The attack scenario and the way it might be exploited.
  • The components involved (CLI, gateway, channels, ClawHub, MCP servers, and so on).
  • Your own severity rating (low / medium / high / critical).
  • References to relevant research, CVEs, or real-world cases.

During review, maintainers handle the ATLAS mapping, threat ID, and risk level.

Offer a mitigation. Open an issue or PR that points to the threat. Be concrete and actionable: "per-sender rate limiting of 10 messages/minute at the gateway" beats "implement rate limiting."

Put forward an attack chain. Attack chains illustrate how several threats combine into a realistic scenario. Lay out the steps and how an attacker would string them together; a brief narrative works better than a rigid template.

Correct or refine existing content. Typos, clarifications, stale details, improved examples: PRs are welcome without opening an issue first.

Framework reference

Threats are linked to MITRE ATLAS (Adversarial Threat Landscape for AI Systems), a framework covering AI/ML-specific threats such as prompt injection, tool misuse, and agent exploitation. Knowing ATLAS is not required to participate; maintainers perform the mapping during review.

Threat IDs. Every threat receives an ID similar to T-EXEC-003, assigned by maintainers during review.

CodeCategory
RECONReconnaissance - information gathering
ACCESSInitial access - gaining entry
EXECExecution - running malicious actions
PERSISTPersistence - maintaining access
EVADEDefense evasion - avoiding detection
DISCDiscovery - learning about the environment
EXFILExfiltration - stealing data
IMPACTImpact - damage or disruption

Risk levels. If the level is unclear to you, simply outline the impact; maintainers will evaluate it.

LevelMeaning
CriticalFull system compromise, or high likelihood + critical impact
HighSignificant damage likely, or medium likelihood + critical impact
MediumModerate risk, or low likelihood + high impact
LowUnlikely and limited impact

Review process

  1. Triage - new submissions are reviewed within 48 hours.
  2. Assessment - maintainers verify feasibility, assign ATLAS mapping and threat ID, validate risk level.
  3. Documentation - formatting and completeness pass.
  4. Merge - added to the threat model and visualization.

Resources

Contact

  • Security vulnerabilities: Trust page for reporting instructions, or security@openclaw.ai.
  • Threat model questions: open a documentation issue.
  • General chat: Discord #security channel.

Recognition

Contributors to the threat model are recognized in the threat model acknowledgments, release notes, and the OpenClaw security hall of fame for significant contributions.

553 words · updated Aug 5, 2026