> For the complete documentation index, see [llms.txt](https://docs.lpp-minduniverse.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lpp-minduniverse.org/lingua-pactum-protocol-lpp-documentation/12.-interoperability-and-standards/12.7-standards-mapping.md). # 12.7 Standards Mapping LPP is not positioned as a replacement for existing identity, authorization, credential, policy, provenance, or evidence standards. The more useful approach is to identify mapping boundaries. ### Verifiable Credentials ``` Verifiable Credentials = portable verifiable claims ``` ``` LPP Authority Objects = action-legitimacy objects evaluated by Layer 0 ``` A VC-compatible representation may potentially carry or support LPP authority claims. No finalized normative mapping is claimed in GitBook v2.0. ### Identity Frameworks ``` Identity Framework ↓ Who is the principal / issuer? ``` ``` LPP ↓ What legitimate authority does that identity possess for this action? ``` Identity therefore supplies evidence into the authority model rather than replacing it. ### Policy Engines ``` Policy Engine = rule evaluation ``` ``` LPP = constitutional admissibility ``` A policy engine may consume Authority Object state. It does not automatically define the origin of that authority. ### Capability Systems ``` Capability = What constrained power can be exercised? ``` ``` LPP = Why should that power be legitimately exercisable for this AI-triggered action? ``` Capabilities may therefore be downstream enforcement artifacts or complements to the Permit model. ### Provenance Systems ``` Provenance System = Where did data / evidence / artifact originate? ``` ``` LPP = How does that evidence support authority, intent, admission, and reconstruction? ``` SourceMind may also integrate with external provenance systems in a full Mind Universe deployment. ### Signed Evidence Systems ``` Signed Evidence System = tamper-evident proof / receipt infrastructure ``` ``` LPP Admission Artifact = governance-specific evidence about an admission decision ``` An external evidence infrastructure may transport, anchor, timestamp, or preserve an LPP Artifact. It does not determine the Layer 0 decision merely by storing the Artifact. ### Authorization Standards OAuth-like, access-token, capability, or authorization mechanisms remain relevant to system access and Layer 1 control. LPP's distinction remains: ``` Authorization ≠ Admissibility ``` The presence of a mature authorization standard does not eliminate the question of upstream action legitimacy. ### Standard Mapping Principle ``` Identity Standard + Credential Standard + Authorization Standard + Policy System + Provenance / Evidence System ↓ Existing Technical Ecosystem LPP ↓ Constitutional Authority + Intent + Admissibility + Execution Evidence Semantics ``` The layers can interoperate without being collapsed. ### Adjacent Work & Emerging Agent Authorization Landscape LPP does not claim that agent identity, delegated authorization, capability attenuation, authorization propagation, or auditable delegation are individually unique primitives. These are active areas of current agent-security research and standards work. The purpose of this comparison is to locate the LPP boundary more precisely rather than to claim an empty competitive field. #### SentinelAgent — Research Paper **SentinelAgent: Intent-Verified Delegation Chains for Securing Federal Multi-Agent AI Systems** formalizes: * verifiable delegation chains, * authority narrowing, * policy preservation, * forensic reconstructibility, * scope-action conformance, * cascade containment, * and probabilistic intent preservation, with enforcement through a non-LLM Delegation Authority Service. The overlap with LPP is substantial around bounded delegation, intent preservation, reconstructability, and runtime enforcement. The LPP positioning difference is that these signals are evaluated inside a broader Layer 0 constitutional admissibility model built around: * externally grounded Authority Objects, * Human-Authorized Intent, * non-compensatory predicates, * constrained Permits, * and Admission Artifacts. #### Authorization Propagation — Research Paper **Authorization Propagation in Multi-Agent AI Systems: Identity Governance as Infrastructure** treats authorization propagation as a workflow-level property and highlights: * transitive delegation, * aggregation inference, * and temporal validity. This work overlaps with LPP's delegation, revocation, and continuity concerns. LPP retains a separate distinction between runtime authorization propagation and the upstream question of whether the propagated authority is constitutionally sufficient for the specific consequential action. #### OIDC-A — Proposed Research Specification **OpenID Connect for Agents (OIDC-A) 1.0** proposes: * agent identity representation, * attestation, * delegation-chain validation, * and capability-based authorization within an OAuth 2.0 / OpenID Connect-compatible model. GitBook v2.0 treats OIDC-A as research / proposal material. It should not be described as an adopted OpenID Foundation specification unless such adoption is separately documented. #### Agent Identity Protocol (AIP) — Individual Internet-Drafts Multiple 2026 individual Internet-Drafts use the AIP name. Examples include drafts focusing on: * verifiable agent identity, * policy enforcement at the tool boundary, * verifiable delegation, * Invocation-Bound Capability Tokens, * and MCP / A2A bindings. These are works in progress. Individual Internet-Drafts are not equivalent to adopted IETF standards. ### Public Positioning Boundary The public novelty claim should therefore remain narrow: > **Identity, delegation, scoped authorization, capability tokens, authorization propagation, and auditable delegation are active shared problem areas.** LPP's positioning claim is the composition of these concerns around a distinct Layer 0 question: > **Does this specific AI-triggered action possess externally grounded authority, preserved human-authorized intent, valid scope and lifecycle state, and sufficient constitutional legitimacy to become an execution candidate at all?** The canonical LPP chain remains: ``` Externally Grounded Authority → Human-Authorized Intent → Constitutional Admissibility → Execution Permit → Enforced Execution Boundary → Reconstructable Evidence ``` This is a positioning boundary, not a claim that every individual component in that chain is unique in isolation. ***