World Standing Together™
Library
WST • PRIVATE • AUDITABLE • HUMAN-FIRST

Peaceful Resolution & Mediation

Modular Continuity Civilization Protocol Reader.

WST · Modular · Continuity · Human-First

Peaceful Resolution & Mediation

Communication and continuity preservation · Foundational Continuity Protocols

Continuity → Recognition → Stewardship → Execution → Settlement → Reconciliation → Verified Continuity
006 — Execution Queue Engine (v1.0).md
/library/_docs/10-governance/006 — Execution Queue Engine (v1.0).md

006 — Execution Queue Engine (v1.0)

WST • PRIVATE • AUDITABLE • HUMAN-FIRST

Purpose

The Execution Queue Engine is the controlled processor that takes accepted intents and advances them into:

transfer.executed

settlement.recorded

fulfillment tracking

It prevents ad hoc execution, bypasses, and silent state changes.

The queue is where agreement stops being potential and becomes system truth.

Core Doctrine

No direct execution from UI

Only READY queue items may execute

Execution must be validated before movement

Execution writes truth

Failures are logged, not hidden

Settlement follows execution, never replaces it

Engine Position in the System

Project

→ intent.created

→ intent.accepted

→ queue item created

→ queue validation

→ transfer.executed

→ settlement hook

→ fulfillment tracking

  1. Queue Purpose

The queue exists to separate:

human declaration

from

system execution

Why this matters:

prevents accidental movement

allows validation before action

supports retries and failure handling

preserves auditability

  1. Canonical Queue Location

/portal/_data/execution/queue.ndjson

Optional per-item drill-down:

/portal/_data/execution/by_item/EXQ-....json

  1. Queue Item Contract

Every queue item should contain:

{

"queue_id": "EXQ-20260417-0001",

"status": "READY",

"intent_id": "INT-20260417-0001",

"project_id": "P-20260416-9NE3A",

"unit_id": "UNIT-001",

"from_member": "WST-M-000002",

"to_member": "WST-M-0000XX",

"charter_no": "WST-CH-US-AR-JONESBORO-001",

"value_class": "service",

"quantity": 1,

"validated": true,

"ready_for_execution": true,

"created_utc": "2026-04-17T15:00:00Z",

"updated_utc": "2026-04-17T15:00:00Z"

}

  1. Queue Status Model

Allowed statuses:

PENDING

READY

EXECUTING

EXECUTED

SETTLING

SETTLED

FULFILLED

FAILED

VOID

Meaning

PENDING: created but not validated

READY: validated and eligible for execution

EXECUTING: engine is actively processing

EXECUTED: transfer.executed written

SETTLING: waiting on settlement evidence or completion

SETTLED: settlement.recorded complete

FULFILLED: real-world completion confirmed

FAILED: processing failed

VOID: item canceled before execution

  1. Queue Entry Rules

A queue item may only be created when:

intent.accepted exists

participants are defined

project/unit reference is valid

value class is known

charter scope is known if applicable

Queue creation event

Recommended internal event:

execution.queued

  1. Validation Rules

Before an item becomes READY, the engine must validate:

Identity

from_member exists

to_member exists if specific target required

Authority

originating seat was valid

charter binding exists where required

Intent

intent.accepted exists

not withdrawn

not already executed

Project linkage

project_id exists

unit_id valid if required

Economic contract

quantity > 0

value_class valid

execution target resolvable

  1. Validation Output

A validated queue item should explicitly show result:

{

"validated": true,

"validation": {

"intent_ok": true,

"authority_ok": true,

"project_ok": true,

"economic_ok": true

},

"ready_for_execution": true

}

If not valid:

{

"validated": false,

"ready_for_execution": false,

"failure_reason": "missing_charter_binding"

}

  1. Execution Trigger Rules

Execution may run:

manually by authorized admin/trustee surface

scheduled by engine poller

automatically when item becomes READY

But always through the engine, never directly from form POST into truth.

  1. Execution Function

Canonical behavior:

READY

→ lock item

→ mark EXECUTING

→ write transfer.executed

→ mark EXECUTED

→ invoke settlement hook

  1. Locking Rule

To avoid double execution, the engine must lock queue items before processing.

Minimum doctrine:

one item, one processor

no concurrent writes to the same queue_id

lock must expire safely if process crashes

Recommended lock file path:

/portal/_data/execution/locks/{queue_id}.lock

  1. Execution Output

When successful, engine writes:

transfer.executed

And updates queue item:

{

"status": "EXECUTED",

"executed_utc": "2026-04-17T15:05:00Z",

"uts_cid": "UTS-....",

"transfer_id": "XFR-...."

}

  1. Settlement Hook

After execution, the engine checks settlement mode.

Internal settlement

If value is internally settled:

write settlement.recorded

mark SETTLED

External settlement

If payment rail is external:

mark SETTLING

wait for verified evidence

then write settlement.recorded

  1. Fulfillment Hook

If the executed item represents real-world work, the engine tracks fulfillment separately.

This may be:

manual confirmation

uploaded proof

milestone record

After confirmation:

fulfillment.confirmed

Queue item becomes:

{

"status": "FULFILLED",

"fulfilled_utc": "2026-04-17T18:20:00Z"

}

  1. Failure Handling

If execution fails:

no partial transfer truth

write failure log

mark item FAILED

preserve reason

allow retry only through controlled path

Example:

{

"status": "FAILED",

"failure_reason": "intent_missing",

"failure_utc": "2026-04-17T15:04:00Z"

}

  1. Retry Rules

Retries are allowed only if:

item is not already EXECUTED

failure reason is resolvable

new validation passes

Retries must increment attempt count:

{

"attempt_count": 2,

"last_retry_utc": "2026-04-17T15:10:00Z"

}

  1. Anti-Duplicate Rules

The engine must refuse execution if:

same intent_id already produced transfer.executed

queue item already marked EXECUTED

same unit already closed if unit is single-use

This prevents ghost replays.

  1. Recommended Internal Files

Queue

/portal/_data/execution/queue.ndjson

Locks

/portal/_data/execution/locks/

Failures / logs

/portal/_data/execution/logs/execution_failures.ndjson

Optional processed archive

/portal/_data/execution/archive/

  1. UI Responsibilities

Project UI may:

create queue candidates

show queue status

show execution result

Project UI may NOT:

directly write transfer.executed

directly write settlement.recorded

That is engine work only.

  1. City Page Responsibilities

City page may summarize:

queued activity

executed activity

fulfilled work

But it is never the control surface for execution.

  1. Minimal Engine Functions

Suggested function set:

wst_execution_queue_add(array $payload): array

Create queue item

wst_execution_queue_validate(array $item): array

Run validation checks

wst_execution_queue_lock(string $queueId): bool

Acquire lock

wst_execution_queue_process(array $item): array

Execute one item

wst_execution_queue_mark_failed(string $queueId, string $reason): void

Record failure

wst_execution_queue_settle(array $item): array

Run settlement hook

wst_execution_queue_fulfill(array $item, array $proof = []): array

Record fulfillment

  1. Minimal Processing Loop

load PENDING/READY items

→ validate

→ mark READY if valid

→ lock first READY item

→ mark EXECUTING

→ write transfer.executed

→ settlement hook

→ unlock

  1. Example Flow for Your Current System

Project:

Jonesboro Water + Power Pilot

Unit:

Survey local water + power needs

Flow:

intent.created

→ intent.accepted

→ execution.queued

→ READY

→ transfer.executed

→ settlement.recorded (if needed)

→ fulfillment.confirmed

That is the first living project-to-economy pipeline.

  1. Anti-Drift Rules

No UI-only completion

No direct transfer write from projects page

No settlement before execution

No execution without queue validation

No retry without reason

No silent failure

  1. Final Principle

The queue is the gate between intention and irreversible truth.

What this unlocks

Once this exists, your system gains:

controlled automation

safe project execution

replay protection

settlement integration

city-level economic reporting