
Security, mindfully.
Understand, assess and design identity and access, from users to AI agents, with the exact source of every rule.
Understand
Tokens, SAML messages and scopes
Paste a token, a SAML message or your OAuth scopes: every finding cites the rule and its source.
Assess
IAM for AI Agents self-assessment
63 controls across 8 domains, tailored to the agents you run. Your answers stay in your browser.
Design
Architecture builder
Draw the access architecture of an AI agent, then check it rule by rule.
IAM and security practitioners
Learn agent identity flows on concrete cases, and check your own tokens and scopes.
CISOs
Measure the IAM maturity of your AI agents, control by control.
Content, tools & accelerators
Security insights and guidance, explained clearly.
Open-source and purpose-built utilities to strengthen your security posture, or help you organize your work.
IdentityJWT Decoder, ExplainedDecode a token and understand every claim in plain words, with an AI agent lens. Runs entirely in your browser.IdentitySAML Decoder, ExplainedDecode a SAMLResponse or an AuthnRequest, with the good practice and its source for every element. Runs entirely in your browser.OAuthScopeNeatTake control over your API & AI Agents OAuth scopes: inventory, duplicates, broad scopes and the risk of each MCP tool. Runs entirely in your browser.ProductivityEisenhower MatrixA productivity app to prioritize tasks by urgency and importance.AnalysisCyberChefEncode, decode, encrypt, and analyse data with 300+ operations, all in your browser.Threat IntelATT&CK NavigatorExplore, annotate, and visualize MITRE ATT&CK matrices to map adversary techniques.
Guided tools to assess and design identity security.
AssessmentIAM for AI Agents Self-AssessmentScore your maturity on 63 controls across 8 domains, tailored to the AI agents you run, in English or French, and get a shareable rosette. Private: answers stay in your browser.ArchitectureArchitecture Builder for AI AccessDesign the access architecture of an AI agent (MCP, Cross-App Access, RAG), then get a deterministic grade where each rule cites its source and its exact strength. In English or French, 100% in your browser.