Skip to content

Dependencies, state and readiness

What a service depends on, what it keeps, how it runs and how far along it is. The other sections are in Contract sections and Configuration and policy.

dependencies

Declares dependencies on other services via their Pacto contracts.

Field Type Required Constraints
name string Yes Non-empty identifier for the dependency
ref string Yes Non-empty. OCI reference (oci://...) or local path (file://... or bare path)
required boolean Yes Whether the dependency is mandatory. Has no default — every entry must set it
compatibility string Yes Non-empty. Valid semver constraint

Note

required is mandatory (no default) — every dependency must declare it. true means the service cannot function without the dependency; false means it degrades gracefully when the dependency is unavailable. It is a declaration of intent that downstream consumers act on (the dashboard uses it to compute blast radius; deployment tooling may use it to gate rollout), not a deployment-time guard Pacto itself enforces.

Dependency reference schemes

Scheme Example Description
oci:// oci://ghcr.io/acme/auth-pacto:1.0.0 OCI registry reference (required for pacto push)
oci:// (no tag) oci://ghcr.io/acme/auth-pacto Resolved to the highest semver tag satisfying compatibility
file:// file://../shared-db Local filesystem path
(bare path) ../shared-db Local filesystem path (shorthand for file://)

When an oci:// reference omits the tag, pacto queries the registry for available tags and selects the highest semver version that satisfies the compatibility constraint. For example, with compatibility: "^2.0.0" and available tags 1.0.0, 2.0.0, 2.3.0, 3.0.0, pacto resolves to 2.3.0. Tag listings are cached in memory for the duration of the command, so multiple dependencies pointing to the same repository only trigger a single registry query.

compatibility selects a version; it does not police one

The constraint is consulted only when the reference omits its tag. A ref that already names a tag or a digest is taken as-is, and the resolved version is never checked back against compatibility. So this validates clean, with no warning:

dependencies:
  - name: auth
    ref: oci://ghcr.io/acme/auth-pacto:1.0.0   # resolves to 1.0.0 …
    required: true
    compatibility: "^2.0.0"                    # … which this does not satisfy

Pinning and constraining are two separate decisions here. If you pin, the pin wins; keep compatibility truthful anyway, because consumers and the lockfile read it as your declared range.

Note

Validation rejects OCI references whose tag or digest is malformed. Tags must follow the OCI tag grammar ([A-Za-z0-9_][A-Za-z0-9._-]{0,127}), and digests must be well-formed (sha256:<64 hex> or sha512:<128 hex>). The same check applies to config and policy refs.

Compatibility constraint examples

Pacto uses Masterminds/semver constraint syntax:

Constraint Matches Use case
^2.0.0 >= 2.0.0, < 3.0.0 Accept patches and minors within a major version
~2.1.0 >= 2.1.0, < 2.2.0 Accept only patches within a minor version
>= 2.0.0 2.0.0 and above (including 3.x, 4.x, …) Track the latest version above a floor
>= 2.0.0, < 4.0.0 2.x and 3.x only Constrain to a range of major versions
* Any version Always resolve to the absolute latest

Warning

Local dependency references (file:// and bare paths) are only allowed during development. pacto push rejects contracts with local dependencies — all refs must use oci:// before publishing.

Practices:

  • Pin by digest in production. oci://...@sha256:...; a tag-based reference produces a validation warning.
  • Give cloud-managed resources a contract. A lightweight Pacto contract for GCP Cloud SQL, AWS SNS or Azure Service Bus, referenced as a dependency, makes that dependency explicit and version-tracked alongside your own.
  • Pin the whole closure with pacto lock. When a pacto.lock file is present, every resolved dependency must match the pinned digest or the command fails. See Lockfile.

workload

Top-level string describing the execution pattern of the workload. Optional. Enum: service, job, scheduled.

Value Description
service A long-running process that serves requests continuously
job A one-shot task that runs to completion and then exits
scheduled A task that runs on a recurring schedule (e.g. cron)
workload: service

state

Top-level section declaring how the service manages state. Optional — a minimal contract (e.g. a lightweight dependency declaration) may omit it entirely. Instead of platforms guessing whether a service needs persistent storage or stable network identity, the contract declares it explicitly.

Field Type Required Enum values
type string Yes stateless, stateful, hybrid
persistence Persistence Yes
dataCriticality string Yes low, medium, high

State types:

Value What it means Example services
stateless No data retained between requests. Any instance can handle any request. Instances are interchangeable. REST APIs, reverse proxies, API gateways
stateful Retains data between requests. Requires stable storage or instance affinity. Databases, message brokers, distributed caches
hybrid Handles requests statelessly but keeps selective in-memory or local state that enriches behavior. Loss of that state degrades but doesn't break the service. APIs with local caches, services with in-memory session stores

How platforms interpret state:

The combination of state.type, persistence.scope and persistence.durability tells a platform exactly what infrastructure a service needs — these are platform-agnostic signals, not Kubernetes prescriptions. See Platform engineers for the full contract-field → platform-decision mapping (Deployment/StatefulSet/PVC and the equivalents on Nomad, ECS or a custom platform).

Data criticality:

Value What it means
low Loss of data has minimal impact. Can be regenerated or is non-essential.
medium Loss has moderate impact. May require manual recovery.
high Loss has severe business impact. Must be prevented. Implies backups, replication, stricter disruption budgets.

Persistence

Field Type Required Enum values
scope string Yes local, shared
durability string Yes ephemeral, persistent
  • local — data is confined to a single instance. Not shared across replicas.
  • shared — data is shared across all instances via a common store.
  • ephemeral — data can be lost on restart without impact. Caches, temp files, reconstructible state.
  • persistent — data must survive restarts. Requires durable storage.

State invariants

Condition Constraint
type: stateless durability must be ephemeral

A stateless service with persistent storage is a contradiction — the JSON Schema catches it.

state:
  type: stateful
  persistence:
    scope: shared
    durability: persistent
  dataCriticality: high

readiness

Optional. A pactoVersion: "2.0" feature. Declares operational readiness state for the service in a provider-neutral way. Each claim has a completion status, optional category and weight. The assessment includes an expiry date and scoring configuration. Pacto computes a readiness score from claim statuses and weights.

Readiness is a declared self-assessment: it is what the service's authors say they have done, and Pacto checks the arithmetic and the expiry, not the underlying work. It is therefore a different question from compliance, which is decided from observed evidence about a running workload. The dashboard keeps them apart: readiness appears on a revision and as a Needs attention category, never as a compliance verdict.

readiness:
  expires: "2027-06-30"  # assessment-level expiry (YYYY-MM-DD)
  minScore: 80           # gate: the derived score must be >= this (omitted ⇒ 100)
  partialCredit: 0.5     # weight multiplier for partial claims (omitted ⇒ 0.5)

  claims:
    - id: dashboard
      type: url
      status: done
      category: observability
      weight: 20
      evidence: https://grafana.example.com/d/service-dashboard
      description: Main production dashboard

    - id: runbook
      type: document
      status: partial
      category: documentation
      weight: 15
      evidence: docs/runbook.md
      description: Incident runbook exists but missing escalation procedures

    - id: security-review
      type: ticket
      status: not-done
      category: security
      weight: 25
      evidence: JIRA-1234
      description: Security assessment pending

    - id: legacy-cleanup
      type: ticket
      status: deferred
      category: code-quality
      weight: 10
      evidence: TECH-5678
      description: Post-launch cleanup task

  history:
    - date: "2026-12-15"
      version: 1.0.0
      author: alice
      description: Initial readiness assessment

readiness.expires, readiness.claims[] and per-claim status are required when readiness is present. readiness.minScore, readiness.partialCredit and readiness.history[] are optional.

readiness fields:

Field Type Required Constraints
expires string Yes Assessment-level expiry as a strict YYYY-MM-DD date. If the current date is past this date, ALL claims earn 0 weight regardless of status. Parsing is exact round-trip: zero-padded fields required and the value must re-serialize unchanged, so 2026-1-1, RFC 3339 timestamps and impossible dates (e.g. 2026-02-30) are rejected.
minScore integer No Gate threshold on the same 0–100 scale as the score. Omitted ⇒ 100 (every claim must be complete). Enforced by pacto validate --readiness and the operator.
partialCredit number No Multiplier for partial status claims (0.0–1.0). Omitted ⇒ 0.5 (half credit).
claims Claim[] Yes At least one claim.
history HistoryEntry[] No Audit trail of readiness assessment changes.

Claim fields

Field Type Required Constraints
id string Yes Stable readiness requirement id. Pattern: ^[a-z0-9]([a-z0-9._-]*[a-z0-9])?$. Unique within the contract. Policies usually target this field.
type string Yes Enum: url, document, ticket, report, artifact, identifier, other. Evidence type.
status string Yes Enum: done, partial, not-done, deferred. Completion state for the claim.
category string No Enum: architecture, testing, code-quality, observability, security, documentation, infrastructure, ci-cd, deployment, resilience, backup-recovery, incident-response, compliance, other. Categorizes the requirement type.
weight integer Yes Contribution to the readiness score. Range 0100.
evidence string Yes Where the evidence lives — a URL, a bundle-relative file path, a ticket ID, anything. Checked for non-emptiness only: Pacto never fetches it, never resolves a path and never verifies the target exists. This is a claim you are making, not one Pacto audits. (Unlike interfaces[].ref and configurations[].schema, which must exist in the bundle.)
description string No Optional human-readable explanation (non-blank when present).

History entry

Field Type Required Constraints
date string Yes Date of the change as YYYY-MM-DD.
version string Yes Contract version at the time of the change. Non-blank.
author string Yes Person or system that made the change. Non-blank.
description string Yes Description of what changed.

The service owner is declared at the contract level, so readiness claims carry no per-claim owner. The readiness score is not stored in the contract — it is computed by tooling (pacto explain, the dashboard and the operator) from claim statuses, weights, the assessment expiry and the current time.

Readiness score and gate

If assessment is expired (today > readiness.expires):
  score = 0

Otherwise:
  earnedWeight = sum of earned weights for all in-scope claims
    - deferred claims: excluded from both numerator and denominator
    - done: full weight
    - partial: round(weight × partialCredit)
    - not-done: 0

  totalWeight = sum of weights for all in-scope claims (excluding deferred)
  score = round(earnedWeight / totalWeight × 100)

passing = score >= minScore          # minScore defaults to 100

Worked example. Using the four claims from the example above (weights 20, 15, 25, 10):

  • legacy-cleanup has status: deferred → excluded from both numerator and denominator
  • In-scope weights: 20 + 15 + 25 = 60 (total)
  • dashboard (done, weight 20) → earns 20
  • runbook (partial, weight 15) → earns round(15 × 0.5) = 8
  • security-review (not-done, weight 25) → earns 0
  • Earned weight: 20 + 8 + 0 = 28
  • Score: round(28 / 60 × 100) = round(46.7) = 47

If the assessment is expired (current date past readiness.expires), the score is 0 regardless of individual claim statuses.

Weights are relative. Only the ratio of weights matters — the score normalizes by totalWeight, so a weight of 20 reads as "20%" only when the weights sum to 100. They can sum to anything; making them sum to 100 just makes each read as a percentage directly. pacto explain and the dashboard show each claim's normalized contribution so you never have to do the math.

Deferred claims are excluded entirely from scoring. Use status: deferred for post-launch cleanup tasks or requirements that don't apply to the current service stage.

The gate (minScore) turns the score from informational into actionable. It is a readiness threshold: with minScore: 80 you require 80% weighted completion; a score below that threshold fails the gate. The gate is evaluated by tooling, not baked into contract validity:

  • pacto explain shows Gate: PASS/FAIL (score N / minScore M).
  • pacto validate --readiness (off by default) fails when score < minScore. It is opt-in because it depends on the current time — making it time-dependent — which would otherwise make plain validate non-deterministic.
  • The operator sets status.readiness.passing and the ReadinessSatisfied condition from the same rule.

Because minScore is an authored literal, a policy can still enforce it org-wide (e.g. require readiness.minScore >= 80) — presence rules stay in policies, the threshold bar lives here.

Enforcing readiness with policies

The base schema never requires a specific claim. Organizational standards are expressed as policies using standard JSON Schema. For example, to require a dashboard claim with status: done, category: observability and weight >= 20:

{
  "type": "object",
  "required": ["readiness"],
  "properties": {
    "readiness": {
      "type": "object",
      "required": ["claims"],
      "properties": {
        "claims": {
          "type": "array",
          "contains": {
            "type": "object",
            "required": ["id", "status", "category", "weight"],
            "properties": {
              "id": { "const": "dashboard" },
              "status": { "const": "done" },
              "category": { "const": "observability" },
              "weight": { "minimum": 20 }
            }
          }
        }
      }
    }
  }
}

Combine multiple contains under allOf to require several claims (e.g. dashboard + runbook + security-review). Constraints JSON Schema cannot express — such as "total weight must equal 100" — are left to a future policy engine.