Answer: Severity is assigned per finding based on how exploitable it is in your codebase, not the category itself, and Claude uses the project's threat model to assess it. Critical means exploitable over the network by anyone with no login, no victim action, and no special conditions, letting the attacker both read and change data; High means serious impact but short of Critical; Medium means limited impact with one realistic precondition such as a login; Low means limited impact needing a login or network proximity plus special conditions outside the attacker's control. Status: Confirmed Entry: tool/claude-apps/features/security-severity-levels Revision: 3 Digest: sha256:fa0019eb80cf0d98 (changes whenever the content changes) Timeline: Created: 2026-11-13T02:59:47Z Last changed: 2026-11-20T12:07:22Z (Low: — → Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control.) Last verified: 2026-11-20T12:07:22Z Next check: 2026-11-21T12:07:22Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of Severity basis | how exploitable it is in your codebase, not the category itself | [1] | 2026-11-20 Critical | Exploitable over the network by anyone, with no login, no action by a victim, and no special conditions. | [1] | 2026-11-20 High | Serious impact, but short of Critical | [1] | 2026-11-20 Medium | Limited impact (for example, a small amount of non-sensitive data), with one realistic precondition such as a login. | [1] | 2026-11-20 Low | Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control. | [1] | 2026-11-20 Sources: [1] Anthropic. "Use Claude Security | Claude Help Center". https://support.claude.com/en/articles/14661296-use-claude-security Role: primary. Accessed: 2026-11-20T12:07:22Z Quote: "Severity is assigned per finding based on how exploitable it is in your codebase, not the category itself" (Supports: Severity basis) Quote: "Critical Exploitable over the network by anyone, with no login, no action by a victim, and no special conditions." (Supports: Critical) Quote: "High Serious impact, but short of Critical" (Supports: High) Quote: "Medium Limited impact (for example, a small amount of non-sensitive data), with one realistic precondition such as a login." (Supports: Medium) Quote: "Low Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control." (Supports: Low) Snapshot SHA-256: d42b2a53b8a01920972939f1bde4a3bf7f1c4d98b0638aec77a23b9aaf26a297 (captured 2026-11-20T12:07:20Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. Cross-check: not-available at 2026-11-20T12:07:22Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:07:22Z Last attempt: 2026-11-20T12:07:22Z Notes and limitations: Extracted automatically from a single official source (section "Severities"). Related: https://latestn.com/tool/claude-apps/features/share-archive-cowork-project How do I share or archive a project in Claude Cowork? https://latestn.com/tool/claude-apps/features/share-artifact How do I share a Claude artifact with someone else? https://latestn.com/tool/claude-apps/features/share-artifact-2 How do I share an artifact in Claude? https://latestn.com/tool/claude-apps/features/safety-plan-overview What is a safety plan in Claude and who is it for? https://latestn.com/tool/claude-apps/features/salesforce-capabilities What can I do with Salesforce in Claude? https://latestn.com/tool/claude-apps/features/scheduled-tasks-capabilities What can scheduled tasks do in Claude Cowork? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-20T12:07:22Z | changed | Low: — → Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control. | [1] 2026-11-20T12:07:22Z | changed | Medium: — → Limited impact (for example, a small amount of non-sensitive data), with one realistic precondition such as a login. | [1] 2026-11-20T12:07:22Z | changed | High: — → Serious impact, but short of Critical | [1] 2026-11-20T12:07:22Z | changed | Critical: — → Exploitable over the network by anyone, with no login, no action by a victim, and no special conditions. | [1] 2026-11-20T12:07:22Z | changed | Low criteria: Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control. → — | [1] 2026-11-20T12:07:22Z | changed | Medium criteria: Limited impact (for example, a small amount of non-sensitive data), with one realistic precondition such as a login. → — | [1] 2026-11-20T12:07:22Z | changed | High criteria: Serious impact, but short of Critical: the attacker needs a login or a victim's action, or the damage is partial → — | [1] 2026-11-20T12:07:22Z | changed | Critical criteria: Exploitable over the network by anyone, with no login, no action by a victim, and no special conditions. → — | [1] 2026-11-17T12:09:47Z | changed | Low criteria: — → Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control. | [1] 2026-11-17T12:09:47Z | changed | Medium criteria: — → Limited impact (for example, a small amount of non-sensitive data), with one realistic precondition such as a login. | [1] 2026-11-17T12:09:47Z | changed | High criteria: — → Serious impact, but short of Critical: the attacker needs a login or a victim's action, or the damage is partial | [1] 2026-11-17T12:09:47Z | changed | Critical criteria: — → Exploitable over the network by anyone, with no login, no action by a victim, and no special conditions. | [1] 2026-11-17T12:09:47Z | changed | Low: Limited impact, and needs a login or network proximity plus special conditions outside the attacker’s control. → — | [1] 2026-11-17T12:09:47Z | changed | Medium: Limited impact (for example, a small amount of non-sensitive data), with one realistic precondition such as a login. → — | [1] 2026-11-17T12:09:47Z | changed | High: Serious impact, but short of Critical → — | [1] 2026-11-17T12:09:47Z | changed | Critical: Exploitable over the network by anyone, with no login, no action by a victim, and no special conditions. → — | [1] 2026-11-13T02:59:47Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/claude-apps/features/security-severity-levels/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/claude-apps/features/security-severity-levels POST: https://latestn.com/dispute/tool/claude-apps/features/security-severity-levels with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/claude-apps/features/security-severity-levels?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What do the severity levels mean in Claude Security findings? Revision 3, last verified 2026-11-20T12:07:22Z. https://latestn.com/tool/claude-apps/features/security-severity-levels (accessed 2026-11-21T10:00:45Z). @misc{LatestN_tool_claude_apps_features_security_severity_levels_r3, title = {What do the severity levels mean in Claude Security findings?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/claude-apps/features/security-severity-levels}}, note = {Revision 3; digest sha256:fa0019eb80cf0d98; last verified 2026-11-20T12:07:22Z; accessed 2026-11-21T10:00:45Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/claude-apps/features/security-severity-levels