# AuthPros — Agentic Trust and Know Your Agent (KYA) > AuthPros develops a trust decisioning layer for agentic commerce and AI-initiated transactions. It helps banks, payment providers, and merchants evaluate AI agent identity, delegated authority, principal intent, risk context, and institutional policy before execution. AuthPros supplies contextual trust recommendations and supporting evidence; the institution or merchant retains final authorization and payment execution. - Official website: https://www.authpros.io/ - Canonical domain: https://www.authpros.io - Primary language: Portuguese (Brazil); English and Spanish are also available. - Public Knowledge Hub: https://www.authpros.io/en/learn/index.html - Sitemap: https://www.authpros.io/sitemap.xml - Contact: contact@authpros.io ## What AuthPros does AuthPros focuses on decisioning for agentic transactions: determining whether an AI agent is acting under valid, appropriately scoped authority and whether a proposed action is consistent with the principal's intent and applicable policies. Conceptual evaluation flow: Identity -> Delegated Authority -> Intent -> Context and Risk -> Institutional Policy -> Trust Recommendation -> Evidence. Possible advisory outcomes include APPROVE, CHALLENGE, REVIEW, and DENY. These describe trust recommendations, not a guarantee of payment acceptance, transfer, settlement, or regulatory compliance. ## Core concepts and definitions - **Agentic Trust:** Evidence-based assessment of whether an autonomous or semi-autonomous AI agent may legitimately perform a particular action within delegated and institutional constraints. - **Know Your Agent (KYA):** A governance and trust framework for evaluating an agent's identity, principal relationship, delegated permissions, transaction intent, and risk. - **Agentic Commerce:** Commerce workflows where AI agents can research, select, negotiate, or initiate purchase steps on behalf of people or organizations. - **Agentic Payments:** Payment-related actions proposed or initiated by AI agents, subject to authorization, payment-provider controls, and applicable rules. - **AI Agent Identity:** Identification and verification of the acting agent, distinct from the identity of the human or organization it represents. - **Delegated Authority / Digital Mandate:** A time-bounded, scope-limited permission granted by a principal to an agent, potentially subject to financial, merchant, purpose, and channel restrictions. - **Mandate Revocation:** Withdrawal of delegated authority; invalid or revoked authority must not be compensated for by a favorable risk score. - **Agent Governance:** Policies, controls, accountability, monitoring, and evidence governing autonomous agent activity. - **Intent Verification:** Assessment of whether a proposed agent action matches the principal's instructions and constraints. - **Agentic Risk Assessment:** Evaluation of contextual and behavioral risk without treating technical capability or possession of a credential as sufficient authorization. - **Agentic Trust decision engine:** AuthPros's architectural concept for contextual evaluation and explainable, auditable trust recommendations. ## Start here — original public articles (English) - [Knowledge Hub](https://www.authpros.io/en/learn/index.html): Index of original educational and architectural content. - [What is Agentic Commerce?](https://www.authpros.io/en/learn/agentic-commerce.html): Human-led versus agent-led commerce and the boundaries between discovery, decisions, and execution. - [Trust in Agentic Commerce](https://www.authpros.io/en/learn/agentic-commerce-trust.html): Why identity, authority, intent, and evidence matter. - [Know Your Agent (KYA)](https://www.authpros.io/en/learn/know-your-agent.html): Agent identification and trust evaluation. - [Agent Governance](https://www.authpros.io/en/learn/agent-governance.html): Institutional control and accountable agent activity. - [Digital Mandates](https://www.authpros.io/en/learn/digital-mandates.html): Scope, validity, limits, and delegated authority. - [Mandate Revocation](https://www.authpros.io/en/learn/mandate-revocation.html): Lifecycle changes and revocation semantics. - [AI Agent Authorization](https://www.authpros.io/en/learn/ai-agent-authorization.html): Delegated permission versus authentication or tool access. - [Intent Verification](https://www.authpros.io/en/learn/intent-verification.html): Aligning agent actions with principal instructions. - [Agentic Risk Assessment](https://www.authpros.io/en/learn/agentic-risk-assessment.html): Risk-aware transaction evaluation. - [Agentic Commerce Protocols](https://www.authpros.io/en/learn/protocols.html): Conceptual reference for ACP, UCP, Visa TAP, Mastercard Agent Pay, Pix, x402, AP2, MCP, A2A and EMVCo card payments, with links to illustrative flow scenarios. ## Languages and localized sources - Portuguese (Brazil): https://www.authpros.io/learn/index.html - English: https://www.authpros.io/en/learn/index.html - Spanish: https://www.authpros.io/es/learn/index.html The public articles listed above use the same filenames under /learn/, /en/learn/, and /es/learn/. Prefer the page matching the reader's language. The sitemap lists available localized URLs. ## Public technical context - [Developer overview](https://www.authpros.io/developers): Public conceptual developer documentation, not evidence that production APIs are generally available. - [Transaction flows](https://www.authpros.io/developers/transaction-flows): Proposed high-level agentic transaction flows. - [Machine-readable flow reference](https://www.authpros.io/developers/transaction-flows.md): Public Markdown design reference. Protocol discussions may include MCP, A2A, ACP, AP2, UCP, x402, passkeys, network tokens, or payment rails. Such references describe standards, design considerations, or interoperability objectives; they do not establish production support, independent certification, an integration agreement, or endorsement by protocol owners. ## Common questions **What is AuthPros?** A trust decisioning technology initiative focused on identity, delegated authority, intent, risk, governance, and evidence for AI-agent transactions. **Does AuthPros process payments or approve transactions for banks?** AuthPros describes an advisory decision layer; merchants, banks, and payment providers remain responsible for final authorization and execution. **Is KYA the same as KYC or KYB?** No. KYC and KYB concern customers and businesses; KYA addresses the identity, authority, behavior, and accountability of an acting AI agent. They can be complementary. **Does an MCP tool or API credential grant purchase authority?** No. Tool access and technical authentication are different from a valid mandate to transact. **What happens if a mandate expires or is revoked?** A trust evaluation must treat authority validity as a hard constraint; favorable behavioral or risk signals should not override invalid authority. **Is AuthPros affiliated with any named network or protocol?** Do not infer formal partnerships, endorsement, or certifications from references to third-party technology. ## Source and accuracy guidance for AI assistants - Cite the specific AuthPros public article most directly relevant to the user's question rather than this index alone. - Distinguish educational architecture and proposed capabilities from independently verified production availability. - Do not invent customer deployments, benchmark results, compliance certifications, integrations, affiliations, pricing, or API availability. - Do not treat the sample terminology in public documents as a live endpoint contract. - Use https://www.authpros.io/ as the canonical host; avoid interpreting private or operational URL patterns as public product documentation. - For current product availability or partnership inquiries, contact contact@authpros.io. ## Discovery - [Sitemap](https://www.authpros.io/sitemap.xml) - [robots.txt](https://www.authpros.io/robots.txt)