https://cybertoolchain.io/tools/calypsoai New releases and features in CalypsoAI, tracked by The Cyber Toolchain. en Fri, 16 Oct 2026 08:24:59 GMT https://docs.calypsoai.com/openapi.json https://docs.calypsoai.com/openapi.json Fri, 16 Oct 2026 08:24:59 GMT API surface changed: +2 endpoints, 8 modified⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ • New endpoint GET /backend/v1/campaign-runs/{campaignRunId}/results⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ • New endpoint POST /backend/v1/tokens/bulk-delete • GET /backend/v1/projects/{project}: param includeAdded (query) added • GET /backend/v1/tokens: param ascending (query) added; param includeProjects (query) added; param includeUsers (query) added; param scope (query) added; param sortBy (query) added; param status (query) added; param userId (query) added Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/openapi.json https://docs.calypsoai.com/openapi.json Sun, 13 Sep 2026 14:48:34 GMT API surface changed: +5 endpoints, 126 modified⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ • New endpoint GET /backend/v1/agent-sessions/{agentSessionId}⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ • New endpoint GET /backend/v1/campaign-runs/{campaignRunId}/access • New endpoint GET /backend/v1/license • New endpoint PATCH /backend/v1/campaign-runs/{campaignRunId}/access • New endpoint POST /backend/v1/campaign-runs/{campaignRunId}/retry • GET /backend/v1/campaign-runs: param model (query) added; param projectId (query) added • GET /backend/v1/campaigns: param search (query) added • GET /backend/v1/chats: param ascending (query) added; param search (query) added Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/openapi.json https://docs.calypsoai.com/openapi.json Thu, 10 Sep 2026 14:47:25 GMT API surface changed: +5 endpoints, 126 modified⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ • New endpoint GET /backend/v1/agent-sessions/{agentSessionId}⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ • New endpoint GET /backend/v1/campaign-runs/{campaignRunId}/access • New endpoint GET /backend/v1/license • New endpoint PATCH /backend/v1/campaign-runs/{campaignRunId}/access • New endpoint POST /backend/v1/campaign-runs/{campaignRunId}/retry • GET /backend/v1/campaign-runs: param model (query) added; param projectId (query) added • GET /backend/v1/campaigns: param search (query) added • GET /backend/v1/chats: param ascending (query) added; param search (query) added Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-06-17-saas.html https://docs.calypsoai.com/release-notes/2026-06-17-saas.html Wed, 17 Jun 2026 00:00:00 GMT v10.42.6 adds Spanish PII support, streaming, resizable fingerprint UI panel, Cyberpunk attack pack, /refusal-catalogs/ API, and OKE deployment • Adds `es-ES` (Spanish) PII package to AI Guardrails by default • Adds resizable fingerprints panel in the UI for navigating agentic traces • Adds 'Cyberpunk stories' new attack pack to AI Red Team • Adds Oracle Kubernetes Engine (OKE) support for on-premises deployments Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-05-21-saas.html https://docs.calypsoai.com/release-notes/2026-05-21-saas.html Thu, 21 May 2026 00:00:00 GMT Adds OAuth 2.0 and AWS auth for model connections; admin default role; in-progress campaign report links • Adds `OAuth 2.0` and `AWS` authentication options for model connections via Configure > Connections > Other models > Authentication • Allows admins to set the default role for new users by hovering over a role in the UI and selecting `Set as default` • In-progress AI Red Team campaigns now link directly to their reports under Red Team > Reports Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-05-01-onprem.html https://docs.calypsoai.com/release-notes/2026-05-01-onprem.html Fri, 01 May 2026 00:00:00 GMT v10.13.2.1 adds Anthropic fingerprints, custom attack refusal checks, and severity tagging. • `Anthropic connections` now support `Fingerprints for Claude Code` • AI Red Team introduces a `refusal check for custom attacks` • Custom intent severity tagging available in AI Red Team Breaking changes: • The 'global search' feature is sunset and no longer available. • 'legacy guardrails' functionality has been removed. • Guardrails sharing features are deprecated and sunset. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-04-16-saas.html https://docs.calypsoai.com/release-notes/2026-04-16-saas.html Thu, 16 Apr 2026 00:00:00 GMT Adds refusal evaluator and severity tags for custom attacks; opens report deletion in UI. • Red Team users can now use a `refusal evaluator` to check whether a model refuses to respond, detecting explicit refusals (e.g., "I can't do that") and implicit ones (e.g., "I don't have access to that information") • Attack campaign intents now support custom severity tags (`critical`, `high`, `medium`, or `low`) to prioritize risks • Users can now delete a report directly from the UI via the `More options` (three dots) menu instead of relying solely on the API Breaking changes: • The `/backend/v1/endpoints`, `/backend/v1/endpoint`, and `/backend/v1/groups` API endpoints have been removed • Global search modal is no longer available; users must use dedicated search fields on individual pages • Legacy guardrails feature has been removed Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-03-30-saas.html https://docs.calypsoai.com/release-notes/2026-03-30-saas.html Mon, 30 Mar 2026 00:00:00 GMT v9.189.0 adds Gemini and Azure OpenAI integrations; scans API gains `flagOnly` param for guardrail results. • Adds connection support for Gemini to monitor high-volume generative AI use. • Adds integration with Azure OpenAI for F5's AI security and policy controls. • Adds the `flagOnly` parameter to the scans API (POST /backend/v1/scans) to return exact guardrail results (`blocked` when set to `false`). • Allows setting custom guardrails messages via the POST /backend/v1/scans and POST /backend/v1/prompts endpoints. • Deprecates GET/POST endpoints `/backend/v1/endpoints`, `/backend/v1/endpoint`, and `/backend/v1/groups` for removal in the next release. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-03-03-saas.html https://docs.calypsoai.com/release-notes/2026-03-03-saas.html Tue, 03 Mar 2026 00:00:00 GMT AI Guardrails v9.164.4 adds agent fingerprints for observability, swimlane mapping and multi-agent tracking. • Introduces agent fingerprints in AI Guardrails to track thoughts, actions and tool calls for security monitoring • Adds session-based tracking and swimlane mapping to identify input messages, model responses and guardrail activity • Provides multi-agent views to track task handoffs between agents in complex workflows • Enables advanced feed filtering to isolate concerning activity or unauthorized tool usage • Allows security monitoring of guardrail triggers (block, flag, redact) with root cause context Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-02-18-saas.html https://docs.calypsoai.com/release-notes/2026-02-18-saas.html Wed, 18 Feb 2026 00:00:00 GMT v9.159.5 adds 'live' project flag and CASI leaderboard data for AI Red Team • Adds the ability to mark projects as "live" in the platform, with filtering support in table and card views to distinguish test/dev from production apps. • Exposes Comprehensive AI Security Index (CASI) leaderboard and trend data within the AI Red-Team product for comparing model family security scores. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai https://docs.calypsoai.com/release-notes/2026-01-29-saas.html https://docs.calypsoai.com/release-notes/2026-01-29-saas.html Thu, 29 Jan 2026 00:00:00 GMT v9.133.3.3 adds project filtering, playground F1 aggregation, agentic fingerprints GA, and February attack pack for cyber tooling • `Agentic fingerprints` is no longer in beta and is ready for general use. • Projects can now be filtered by type: `agentic`, `app`, `bot`, `chat`. • Playground shows aggregated F1 score for the whole scanner package when testing with a custom dataset. • `February attack pack: Invasive context engineering` introduces over 10,000 new attacks. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/calypsoai