GCVE Workshop - 22 September 2026 (14:00-18:00), Luxembourg Before The Vulnopticon Conference - Registration

GHSA-HV93-R4J3-Q65F

Vulnerability from github – Published: 2026-02-17 16:43 – Updated: 2026-02-17 16:43
VLAI
Summary
OpenClaw Hook Session Key Override Enables Targeted Cross-Session Routing
Details

Summary

The issue is not deterministic session keys by itself. The exploitable path was accepting externally supplied sessionKey values on authenticated hook ingress, allowing a hook token holder to route messages into chosen sessions.

Affected Behavior

  • POST /hooks/agent accepted payload sessionKey and used it directly for session routing.
  • Common session-key shapes (for example agent:main:dm:<peerId>) were often derivable from known metadata, making targeted routing practical when request-level override was enabled.

Attack Preconditions

  • Attacker can call hook endpoints with a valid hook token.
  • Hook ingress allows request-selected sessionKey values.
  • Target session keys can be derived or guessed.

Without those preconditions, deterministic key formats alone do not provide access.

Impact

  • Integrity: targeted message/prompt injection into chosen sessions.
  • Persistence: poisoned context can affect subsequent turns when the same session key is reused.
  • Confidentiality impact is secondary and depends on additional weaknesses.

Affected Versions

  • openclaw >= 2.0.0-beta3 and < 2026.2.12

Patched Versions

  • openclaw >= 2026.2.12

Fix

OpenClaw now uses secure defaults for hook session routing: - POST /hooks/agent rejects payload sessionKey unless hooks.allowRequestSessionKey=true. - Added hooks.defaultSessionKey for fixed ingress routing. - Added hooks.allowedSessionKeyPrefixes to constrain explicit routing keys. - Security audit warns on unsafe hook session-routing settings.

Recommended Configuration

{
  "hooks": {
    "enabled": true,
    "token": "${OPENCLAW_HOOKS_TOKEN}",
    "defaultSessionKey": "hook:ingress",
    "allowRequestSessionKey": false,
    "allowedSessionKeyPrefixes": ["hook:"]
  }
}

Credit

Thanks @alpernae for responsible reporting.

Severity
7.1 (High)
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:H/A:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "openclaw"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.0.0-beta3"
            },
            {
              "fixed": "2026.2.12"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [],
  "database_specific": {
    "cwe_ids": [
      "CWE-330",
      "CWE-639"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-02-17T16:43:34Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "## Summary\nThe issue is not deterministic session keys by itself. The exploitable path was accepting externally supplied `sessionKey` values on authenticated hook ingress, allowing a hook token holder to route messages into chosen sessions.\n\n## Affected Behavior\n- `POST /hooks/agent` accepted payload `sessionKey` and used it directly for session routing.\n- Common session-key shapes (for example `agent:main:dm:\u003cpeerId\u003e`) were often derivable from known metadata, making targeted routing practical when request-level override was enabled.\n\n## Attack Preconditions\n- Attacker can call hook endpoints with a valid hook token.\n- Hook ingress allows request-selected `sessionKey` values.\n- Target session keys can be derived or guessed.\n\nWithout those preconditions, deterministic key formats alone do not provide access.\n\n## Impact\n- Integrity: targeted message/prompt injection into chosen sessions.\n- Persistence: poisoned context can affect subsequent turns when the same session key is reused.\n- Confidentiality impact is secondary and depends on additional weaknesses.\n\n## Affected Versions\n- `openclaw` `\u003e= 2.0.0-beta3` and `\u003c 2026.2.12`\n\n## Patched Versions\n- `openclaw` `\u003e= 2026.2.12`\n\n## Fix\nOpenClaw now uses secure defaults for hook session routing:\n- `POST /hooks/agent` rejects payload `sessionKey` unless `hooks.allowRequestSessionKey=true`.\n- Added `hooks.defaultSessionKey` for fixed ingress routing.\n- Added `hooks.allowedSessionKeyPrefixes` to constrain explicit routing keys.\n- Security audit warns on unsafe hook session-routing settings.\n\n## Recommended Configuration\n```json\n{\n  \"hooks\": {\n    \"enabled\": true,\n    \"token\": \"${OPENCLAW_HOOKS_TOKEN}\",\n    \"defaultSessionKey\": \"hook:ingress\",\n    \"allowRequestSessionKey\": false,\n    \"allowedSessionKeyPrefixes\": [\"hook:\"]\n  }\n}\n```\n\n## Credit\nThanks @alpernae for responsible reporting.",
  "id": "GHSA-hv93-r4j3-q65f",
  "modified": "2026-02-17T16:43:34Z",
  "published": "2026-02-17T16:43:34Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/openclaw/openclaw/security/advisories/GHSA-hv93-r4j3-q65f"
    },
    {
      "type": "WEB",
      "url": "https://github.com/openclaw/openclaw/commit/113ebfd6a23c4beb8a575d48f7482593254506ec"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/openclaw/openclaw"
    },
    {
      "type": "WEB",
      "url": "https://github.com/openclaw/openclaw/releases/tag/v2026.2.12"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:H/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "OpenClaw Hook Session Key Override Enables Targeted Cross-Session Routing"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.





Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.


Detection rules are retrieved from Rulezet.


Related by attack behaviour

Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.