{"id":"GHSA-56f2-hvwg-5743","aliases":[],"status":"published","title":"OpenClaw affected by SSRF in Image Tool Remote Fetch","description":"## Summary\n\nA server-side request forgery (SSRF) vulnerability in the Image tool allowed attackers to force OpenClaw to make HTTP requests to arbitrary internal or restricted network targets.\n\n## Affected Versions\n\n- npm: openclaw \u003c= 2026.2.1\n\n## Patched Versions\n\n- npm: openclaw 2026.2.2 and later\n\n## Fix Commits\n\n- 81c68f582d4a9a20d9cca9f367d2da9edc5a65ae (guard remote media fetches with SSRF checks)\n- 9bd64c8a1f91dda602afc1d5246a2ff2be164647 (expand SSRF guard coverage)\n\n## Details\n\nThe Image tool accepts file paths, file:// URLs, data: URLs, and http(s) URLs. In vulnerable versions, http(s) URLs were fetched without SSRF protections, enabling requests to localhost, RFC1918, link-local, and cloud metadata targets.\n\nThis was fixed by routing remote media fetching through the SSRF guard (private/internal IP + hostname blocking, redirect hardening, DNS pinning).\n\n## Exploitability Notes\n\n- Requires attacker-controlled invocation of the Image tool (direct tool access, or a gateway/channel surface that forwards untrusted `image` arguments into tool calls).\n- The image tool expects the fetched content to be an image. Many high-value SSRF targets return text/JSON (for example cloud metadata endpoints), which will typically fail media-type validation. In practice, the most direct confidentiality impact comes from internal endpoints that actually return images (screenshots/renderers, camera snapshots, chart exports, etc.).\n- Remote fetches are GET-only with no custom headers. Some metadata services require special headers or session tokens (for example GCP `Metadata-Flavor`, AWS IMDSv2 token), which can further reduce the likelihood of direct credential theft in some environments.\n- Despite the above constraints, SSRF remains a powerful primitive: it can enable internal network probing and access to unauthenticated/internal HTTP endpoints, and can chain with other weaknesses if present.\n\nThanks @p80n-sec for reporting.","severity":{"version":"3.1","score":7.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:L","level":"high","from":"ghsa"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":true},"cwes":["CWE-918"],"published":"2026-02-17T17:13:35Z","modified":"2026-02-22T23:31:45Z","fixes":[{"ecosystem":"npm","package":"openclaw","fixed_in":["2026.2.2"],"sources":["ghsa","osv"]}],"affected":[{"source":"ghsa","ecosystem":"npm","package":"openclaw","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"fixed":"2026.2.2"}]}]},{"source":"osv","ecosystem":"npm","package":"openclaw","ranges":[{"type":"SEMVER","events":[{"introduced":"0"},{"fixed":"2026.2.2"}]}]}],"references":[{"url":"https://github.com/openclaw/openclaw/security/advisories/GHSA-56f2-hvwg-5743","tags":["web"]},{"url":"https://github.com/openclaw/openclaw/commit/81c68f582d4a9a20d9cca9f367d2da9edc5a65ae","tags":["web"]},{"url":"https://github.com/openclaw/openclaw/commit/9bd64c8a1f91dda602afc1d5246a2ff2be164647","tags":["web"]},{"url":"https://github.com/openclaw/openclaw/releases/tag/v2026.2.2","tags":["web"]},{"url":"https://github.com/advisories/GHSA-56f2-hvwg-5743"},{"url":"https://github.com/openclaw/openclaw","tags":["package"]}],"sources":{"ghsa":[{"id":"GHSA-56f2-hvwg-5743","title":"OpenClaw affected by SSRF in Image Tool Remote Fetch","description":"## Summary\n\nA server-side request forgery (SSRF) vulnerability in the Image tool allowed attackers to force OpenClaw to make HTTP requests to arbitrary internal or restricted network targets.\n\n## Affected Versions\n\n- npm: openclaw \u003c= 2026.2.1\n\n## Patched Versions\n\n- npm: openclaw 2026.2.2 and later\n\n## Fix Commits\n\n- 81c68f582d4a9a20d9cca9f367d2da9edc5a65ae (guard remote media fetches with SSRF checks)\n- 9bd64c8a1f91dda602afc1d5246a2ff2be164647 (expand SSRF guard coverage)\n\n## Details\n\nThe Image tool accepts file paths, file:// URLs, data: URLs, and http(s) URLs. In vulnerable versions, http(s) URLs were fetched without SSRF protections, enabling requests to localhost, RFC1918, link-local, and cloud metadata targets.\n\nThis was fixed by routing remote media fetching through the SSRF guard (private/internal IP + hostname blocking, redirect hardening, DNS pinning).\n\n## Exploitability Notes\n\n- Requires attacker-controlled invocation of the Image tool (direct tool access, or a gateway/channel surface that forwards untrusted `image` arguments into tool calls).\n- The image tool expects the fetched content to be an image. Many high-value SSRF targets return text/JSON (for example cloud metadata endpoints), which will typically fail media-type validation. In practice, the most direct confidentiality impact comes from internal endpoints that actually return images (screenshots/renderers, camera snapshots, chart exports, etc.).\n- Remote fetches are GET-only with no custom headers. Some metadata services require special headers or session tokens (for example GCP `Metadata-Flavor`, AWS IMDSv2 token), which can further reduce the likelihood of direct credential theft in some environments.\n- Despite the above constraints, SSRF remains a powerful primitive: it can enable internal network probing and access to unauthenticated/internal HTTP endpoints, and can chain with other weaknesses if present.\n\nThanks @p80n-sec for reporting.","status":"published","published":"2026-02-17T17:13:35Z","modified":"2026-02-17T17:13:36Z","cvss":[{"version":"3.1","score":7.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:L","level":"high"}],"cwes":["CWE-918"],"updated_at":"2026-11-02T14:39:25.82905Z"}],"osv":[{"id":"GHSA-56f2-hvwg-5743","title":"OpenClaw affected by SSRF in Image Tool Remote Fetch","description":"## Summary\n\nA server-side request forgery (SSRF) vulnerability in the Image tool allowed attackers to force OpenClaw to make HTTP requests to arbitrary internal or restricted network targets.\n\n## Affected Versions\n\n- npm: openclaw \u003c= 2026.2.1\n\n## Patched Versions\n\n- npm: openclaw 2026.2.2 and later\n\n## Fix Commits\n\n- 81c68f582d4a9a20d9cca9f367d2da9edc5a65ae (guard remote media fetches with SSRF checks)\n- 9bd64c8a1f91dda602afc1d5246a2ff2be164647 (expand SSRF guard coverage)\n\n## Details\n\nThe Image tool accepts file paths, file:// URLs, data: URLs, and http(s) URLs. In vulnerable versions, http(s) URLs were fetched without SSRF protections, enabling requests to localhost, RFC1918, link-local, and cloud metadata targets.\n\nThis was fixed by routing remote media fetching through the SSRF guard (private/internal IP + hostname blocking, redirect hardening, DNS pinning).\n\n## Exploitability Notes\n\n- Requires attacker-controlled invocation of the Image tool (direct tool access, or a gateway/channel surface that forwards untrusted `image` arguments into tool calls).\n- The image tool expects the fetched content to be an image. Many high-value SSRF targets return text/JSON (for example cloud metadata endpoints), which will typically fail media-type validation. In practice, the most direct confidentiality impact comes from internal endpoints that actually return images (screenshots/renderers, camera snapshots, chart exports, etc.).\n- Remote fetches are GET-only with no custom headers. Some metadata services require special headers or session tokens (for example GCP `Metadata-Flavor`, AWS IMDSv2 token), which can further reduce the likelihood of direct credential theft in some environments.\n- Despite the above constraints, SSRF remains a powerful primitive: it can enable internal network probing and access to unauthenticated/internal HTTP endpoints, and can chain with other weaknesses if present.\n\nThanks @p80n-sec for reporting.","status":"published","published":"2026-02-17T17:13:35Z","modified":"2026-02-22T23:31:45.590805Z","cvss":[{"version":"3.1","score":7.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:L","level":"high"}],"cwes":["CWE-918"],"updated_at":"2026-11-02T15:46:37.864461Z"}]}}