SILT Core
Open semantic infrastructure for modelling authority, capacity, consent, delegation, and revocation in digital systems.
SILT Core v0.1 – Semantic Layer Definition
Initial public specification of SILT Core, defining capacity, authority, consent, and revocation as preconditions for legitimate digital action.
SILT Core is designed as an open, implementation-neutral semantic layer. It does not replace DID, VC, zCap, wallet, DAO, or AI-agent frameworks. It defines authority semantics that such systems may adopt, reference, or extend.
Systems can satisfy every authorisation check and still enable illegitimate action, because capacity and authority are not expressed at the point of execution.
Where SILT sits
Most digital systems follow a simple structure:
- Identity → who is acting
- Credentials → what can be proven
- Policy → what is allowed
SILT introduces a missing layer between credentials and policy:
- Capacity → how the actor is acting
- Authority → under what mandate
- Consent → under what conditions
- Revocation → whether that authority still holds
SILT does not decide what actions are permitted. It determines whether an actor is in a position to act at all, under valid authority.
Why this matters
Digital systems are increasingly capable of executing actions automatically through APIs, smart contracts, and AI agents.
Without explicit models of capacity and authority, these systems rely on implicit assumptions about who is allowed to act and why.
SILT makes those conditions explicit, so that actions can be evaluated not only for correctness, but for legitimacy.
What it is
SILT Core is a specification-first framework for modelling authority, capacity, consent, delegation, reliance, and revocation in digital systems.
Rather than delivering an application, wallet, or platform, SILT Core defines a semantic layer that many civic, governance, commercial, and agentic systems implicitly depend on but rarely specify: the conditions under which an actor is authorised to act, bind, delegate, rely, or revoke.
SILT Core draws on long-standing concepts from agency, mandate, private law, governance, and digital identity, but expresses them as technology-agnostic primitives suitable for modern systems.
The current work establishes specifications, threat models, misuse cases, and minimal validation schemas. Implementation is intentionally deferred so the core grammar can remain open, portable, and interoperable.
Interoperability
SILT Core is designed to complement existing identity and coordination frameworks rather than replace them.
DID and VC systems can express identifiers and attestations. Capability systems such as zCaps can express delegated permissions. Wallets, DAOs, and AI-agent frameworks can execute actions.
SILT Core focuses on the missing semantic question beneath these systems:
In what capacity is this actor acting, under what authority, within what scope, and how can that authority be revoked?
This makes SILT Core relevant to DID/VC, ZK, ENS, DAO governance, AI-agent systems, and institutional workflows without requiring any single implementation path.
Why This Matters for Bitcoin
Bitcoin transactions are acts of authority, not just technical events.
Each signature expresses a claim of capacity to act and transfer value. In legal terms this resembles long-standing principles from merchant law, where the validity of an instrument depends not only on the signature itself but on the authority of the signer to bind a person, organisation, or trust.
Every signature expresses intent, responsibility, and consequence. Yet most Bitcoin wallets treat all signatures as equivalent, collapsing different forms of authority into a single approval moment.
In practice, people act in different capacities – personally, as delegates, as stewards, or under temporary or limited authority. When these distinctions are invisible, users make irreversible mistakes, lose funds, or unintentionally bind themselves in ways they did not understand.
SILT Core addresses this gap by specifying primitives for capacity, consent, delegation, and revocation. These primitives make authority explicit and auditable without changing Bitcoin itself.
This work is not a wallet, protocol, or identity system. It is a design and specification layer that Bitcoin tools can adopt to improve signing clarity, reduce catastrophic error, and better align wallet behaviour with how humans actually understand responsibility.
Why This Matters for Ethereum
Ethereum introduced programmable coordination. Smart contracts allow organisations, treasuries, and governance systems to operate without central platforms.
Yet most Ethereum systems still rely on simplified identity assumptions. A wallet address signs a transaction, a multisig controls funds, or a token represents voting power.
These mechanisms execute rules, but they rarely model the deeper semantics of authority.
A signature does not tell us in what capacity an actor is acting, whether authority is delegated or original, what scope that delegation has, when that authority expires, or how it may be revoked.
SILT introduces a semantic layer that makes authority explicit. It formalises capacity, mandate, consent, and revocation so that digital systems can model how an actor is authorised to act, not just whether a key signed a transaction.
Ethereum provides programmable execution. SILT provides the authority grammar that makes execution legitimate.
Why This Matters for AI Systems
Modern AI systems increasingly operate as delegated actors within digital environments. They schedule actions, execute transactions, manage information flows, and interact with institutional systems on behalf of humans or organisations.
As these systems become more agentic, the central question is no longer only what they can do, but under what authority they are permitted to act.
Most digital infrastructure can represent identities, accounts, or permissions, yet rarely models the deeper structure of lawful action: capacity, mandate, consent, delegation, and revocation.
SILT Core explores whether these authority relationships can be expressed as machine-readable primitives. By making acting capacity and delegation explicit, digital systems may support safer forms of automated and AI-mediated action.
What it is not
- Not a wallet, identity app, or credential issuer
- Not a blockchain protocol or chain-specific framework
- Not a token model, DAO toolkit, or governance platform
- Not a replacement for existing identity standards
Current status
- Specifications and schemas: v0.1 released as a versioned baseline
- Threat model and misuse cases: published and embedded in the core model
- Interoperability: exploratory mapping toward DID/VC, capability systems, DAOs, and AI-agent frameworks
- Implementation: intentionally deferred; iteration through external adoption
Open infrastructure
SILT Core is developed as open semantic infrastructure. Core specifications, schemas, and threat models are intended to remain publicly accessible so they can be adopted, tested, criticised, and extended across ecosystems.
Downstream implementations, assurance methods, certification layers, and high-integrity validation services may evolve separately from the open core.
Links
SILT Core is currently being developed as an open specification-first framework. Further work is underway to refine the semantic layer, expand interoperability mappings, and explore downstream implementation pathways.