# Protocol 0.1 — Schemas, evidence and integration | HelmLoop

> HelmLoop makes authority, evidence, promotion and observed outcomes explicit across AI agent workflows. Explore the executable 0.1 candidate and the proposed 0.2 protocol.

Source: https://helmloop.dev/engineering.html

Protocol 0.1 includes eight record schemas, a deterministic six-axis reducer, a CLI and two fixture observation adapters. These are local implementation artifacts, not a public package release or live-runtime certification.

The 0.2 plan adds an assurance axis, authority lifecycle events, criterion-bound outcome observations and versioned evidence admission. These changes are proposed in r3 / sync-4.

## Eight executable 0.1 records

- [OutcomeContractV1](https://helmloop.dev/protocol/v0.1/schemas/outcome-contract-v1.schema.json): Accountable outcome, risk, owner, horizon, criteria, constraints, preferences, stop conditions, and all six axes.
- [AuthorityGrantV1](https://helmloop.dev/protocol/v0.1/schemas/authority-grant-v1.schema.json): Owner-issued, exact, expiring, revocable, single-use authority with durable reservation and terminal lineage.
- [ExecutionReceiptV1](https://helmloop.dev/protocol/v0.1/schemas/execution-receipt-v1.schema.json): Attempt-bound execution fact with resource owner, provider-request binding, and exact reserved authority reference.
- [EvidenceRecordV1](https://helmloop.dev/protocol/v0.1/schemas/evidence-record-v1.schema.json): Freshness-capped, criterion-digested conclusion with provenance, dependency fingerprints, and limitations.
- [AssuranceDecisionV1](https://helmloop.dev/protocol/v0.1/schemas/assurance-decision-v1.schema.json): Evidence-bound L0–L3 assurance decision whose sufficiency is recomputed.
- [PromotionDecisionV1](https://helmloop.dev/protocol/v0.1/schemas/promotion-decision-v1.schema.json): Owner decision with exact target parameters, separate promote authority, and non-regressing lineage.
- [ReconciliationRecordV1](https://helmloop.dev/protocol/v0.1/schemas/reconciliation-record-v1.schema.json): Fresh query bound to the exact execution, attempt, provider request, locator, and observed external state.
- [LearningSignalV1](https://helmloop.dev/protocol/v0.1/schemas/learning-signal-v1.schema.json): Bounded proposal based on outcome evidence; never self-accepting and never truth by itself.

## Six current decision axes

- execution: What the executor did
- evidence: Whether required claims are fresh and supported
- authority: Whether the exact action was authorized
- promotion: What the accountable owner decided
- reconciliation: What the provider and observed external state show
- observedOutcome: What independent outcome evidence shows

0.1 keeps HL-C14N-1, its original readers and six-axis semantics. Planned 0.2 uses helmloop-jcs-sha256-v1 and typed sha256: references. A matching digest checks integrity; it does not grant trust or authority.

## Read the evidence at its actual scope

Current fixture results do not certify a live runtime, broad portability, production readiness or business value. The research page keeps these open questions visible.