Execution API

The Scrums.com API is planned and not yet publicly available. Endpoints and behaviour are subject to change before release.

Overview

The Execution API provides a unified view of all active and historical execution across Service Lines. Regardless of whether execution is running through a talent engagement, a managed delivery pod, an ODS outcome cycle, infrastructure operations, or an agent runtime, this API surfaces operational state consistently under a single interface. The Execution API is a read and management layer. Work is initiated through product line APIs (talent requests, ODS scopes, agent runs). Execution records reflect the operational state of each active LIN-*.

Core Concepts

Execution records

Every LIN-* in active or paused status has a corresponding execution record. Execution records are created automatically when a Service Line activates and closed when it terminates.

Execution states

State	Description
`running`	Work is actively in progress
`idle`	Line is active but no current activity
`paused`	Temporarily suspended; can resume
`blocked`	Action required to proceed (approval gap, capacity issue, policy violation)
`closed`	Execution has ended

Runs

For execution models that have discrete invocations (agent_runtime, timeboxed_outcome), execution is tracked as individual runs (RUN-*). For continuous models (capacity, ongoing_sla, infra_managed), execution is tracked as a continuous record with periodic check-ins.

Execution models and their records

Execution model	Record type	Run identifier
`capacity`	Continuous	Not applicable
`ongoing_sla`	Continuous	Not applicable
`timeboxed_outcome`	Cycle-based	`RUN-*` per ODS cycle
`infra_managed`	Continuous	Not applicable
`observability`	Continuous	Not applicable
`agent_runtime`	Invocation-based	`RUN-*` per agent run

Endpoints

GET /v1/execution

List all execution records for the authenticated context.

Request

GET /v1/execution?workspace_id=WS-26-000021&state=running
Authorization: Bearer <token>

Query parameters

Parameter	Type	Description
`workspace_id`	string	Filter by workspace
`state`	enum	`running`, `idle`, `paused`, `blocked`, `closed`
`execution_model`	enum	Filter by execution model
`product_line`	enum	Filter by product line
`project_tag`	string	Filter by project label
`cursor`	string	Pagination cursor
`limit`	integer	Results per page (max 100, default 25)

Response

{
  "data": [
    {
      "id": "RUN-26-000419",
      "service_line_id": "LIN-26-084729",
      "execution_model": "capacity",
      "state": "running",
      "workspace_id": "WS-26-000021",
      "product_line": "talent",
      "project_tags": ["PROJ-26-004281"],
      "started_at": "2026-01-15T00:00:00Z",
      "last_activity_at": "2026-04-15T08:30:00Z",
      "metrics": {
        "hours_this_week": 32.0,
        "hours_total": 1248.5,
        "health_score": 94
      }
    },
    {
      "id": "RUN-26-000388",
      "service_line_id": "LIN-26-091100",
      "execution_model": "agent_runtime",
      "state": "running",
      "workspace_id": "WS-26-000021",
      "product_line": "agent_gateway",
      "project_tags": [],
      "started_at": "2026-04-10T12:00:00Z",
      "last_activity_at": "2026-04-15T09:12:44Z",
      "metrics": {
        "invocations_today": 47,
        "invocations_total": 1203,
        "last_run_duration_ms": 2411
      }
    }
  ],
  "meta": {
    "cursor": "eyJpZCI6IlJVTi0yNi0wMDA0MTkifQ",
    "has_more": false
  }
}

GET /v1/execution/

Retrieve the execution record for a specific Service Line.

Request

GET /v1/execution/LIN-26-084729
Authorization: Bearer <token>

Response

{
  "data": {
    "service_line_id": "LIN-26-084729",
    "execution_model": "capacity",
    "state": "running",
    "workspace_id": "WS-26-000021",
    "product_line": "talent",
    "project_tags": ["PROJ-26-004281"],
    "resources": [
      {
        "type": "talent",
        "id": "SCR-128491",
        "name": "Allocated engineer",
        "role": "senior_backend_engineer",
        "allocated_hours_per_week": 40,
        "status": "active"
      }
    ],
    "health": {
      "score": 94,
      "signals": [
        { "signal": "hours_on_target", "status": "green" },
        { "signal": "sla_compliance", "status": "green" },
        { "signal": "capacity_utilisation", "status": "amber" }
      ]
    },
    "started_at": "2026-01-15T00:00:00Z",
    "last_activity_at": "2026-04-15T08:30:00Z"
  }
}

GET /v1/execution//runs

List discrete runs for execution models that support them (agent_runtime, timeboxed_outcome).

Request

GET /v1/execution/LIN-26-091100/runs?limit=10
Authorization: Bearer <token>

Response

{
  "data": [
    {
      "id": "RUN-26-001812",
      "service_line_id": "LIN-26-091100",
      "status": "completed",
      "triggered_by": {
        "type": "api",
        "actor_id": "USR-26-000044"
      },
      "input": {
        "prompt": "Summarise open tasks for project PROJ-26-004281"
      },
      "output": {
        "summary": "14 open tasks across 3 sprints. 2 are blocked.",
        "artifact_id": "ART-26-000991"
      },
      "duration_ms": 2411,
      "tokens_used": 1840,
      "started_at": "2026-04-15T09:12:44Z",
      "completed_at": "2026-04-15T09:12:46Z"
    }
  ],
  "meta": {
    "cursor": "eyJpZCI6IlJVTi0yNi0wMDE4MTIifQ",
    "has_more": true
  }
}

Notes

Only available for agent_runtime and timeboxed_outcome Service Lines.
Continuous execution models (capacity, ongoing_sla, infra_managed) do not produce discrete run records.

POST /v1/execution//pause

Pause an active execution. The Service Line remains active but work stops.

Request

{
  "reason": "Stakeholder review period - resuming 2026-05-01"
}

POST /v1/execution//resume

Resume a paused execution.

Common Workflows

Cross-platform execution dashboard

# All running Service Lines across a workspace
GET /v1/execution?workspace_id=WS-26-000021&state=running

# Drill into any blocked lines
GET /v1/execution?workspace_id=WS-26-000021&state=blocked

# Get detail on a specific line
GET /v1/execution/LIN-26-084729

Checking execution health before invoicing

# Before generating a monthly report, confirm all active lines are healthy
GET /v1/execution?workspace_id=WS-26-000021&state=running

# Review health signals for any amber/red lines
GET /v1/execution/LIN-26-084729

Objects

Execution record

Field	Type	Description
`service_line_id`	string	Linked `LIN-*` identifier
`execution_model`	enum	Execution model of the Service Line
`state`	enum	`running`, `idle`, `paused`, `blocked`, `closed`
`workspace_id`	string	Parent workspace
`product_line`	enum	Product line driving execution
`project_tags`	array	Inherited from the Service Line
`resources`	array	Allocated resources contributing to execution
`health`	object	Health score and individual signal statuses
`started_at`	datetime	When execution began
`last_activity_at`	datetime	Most recent activity timestamp

Run record (`RUN-*`)

Field	Type	Description
`id`	string	`RUN-*` identifier
`service_line_id`	string	Linked `LIN-*` identifier
`status`	enum	`pending`, `running`, `completed`, `failed`, `cancelled`
`triggered_by`	object	Actor type and ID
`input`	object	Run input payload
`output`	object	Run output payload (populated on completion)
`duration_ms`	integer	Execution duration in milliseconds
`started_at`	datetime	Run start timestamp
`completed_at`	datetime	Run completion timestamp

Best Practices

Poll execution state, not individual product line APIs. The Execution API gives a consistent state view regardless of product line. Build operational dashboards here, not against talent, ODS, or agent-specific endpoints.
Monitor blocked states proactively. A blocked execution record requires human or system action. Subscribe to execution.blocked events via webhooks to catch these immediately.
Use run records for agent audit trails. For agent_runtime Service Lines, run records provide the complete execution history including inputs, outputs, and durations. This data feeds Audit Logs.

Overview

Organizations

Workspaces

Service Lines

Core Platform

Observability & Metrics

ODS

Talent Marketplace

AI Agent Gateway

Product Lines

CLI

Overview

Core Concepts

Execution records

Execution states

Runs

Execution models and their records

Endpoints

GET /v1/execution

Request

Response

GET /v1/execution/

Request

Response

GET /v1/execution//runs

Request

Response

Notes

POST /v1/execution//pause

Request

POST /v1/execution//resume

Common Workflows

Cross-platform execution dashboard

Checking execution health before invoicing

Objects

Execution record

Run record (`RUN-*`)

Best Practices

​Overview

​Core Concepts

​Execution records

​Execution states

​Runs

​Execution models and their records

​Endpoints

​GET /v1/execution

​Request

​Response

​GET /v1/execution/

​Request

​Response

​GET /v1/execution//runs

​Request

​Response

​Notes

​POST /v1/execution//pause

​Request

​POST /v1/execution//resume

​Common Workflows

​Cross-platform execution dashboard

​Checking execution health before invoicing

​Objects

​Execution record

​Run record (RUN-*)

​Best Practices

Overview

Core Concepts

Execution records

Execution states

Runs

Execution models and their records

Endpoints

GET /v1/execution

Request

Response

GET /v1/execution/

Request

Response

GET /v1/execution//runs

Request

Response

Notes

POST /v1/execution//pause

Request

POST /v1/execution//resume

Common Workflows

Cross-platform execution dashboard

Checking execution health before invoicing

Objects

Execution record

Run record (`RUN-*`)

Best Practices