Technical Schemas
import { Tabs, TabItem } from ‘@astrojs/starlight/components’;
This reference covers the structured JSON artifacts produced and consumed by the context pipeline.
Architect output schema
Section titled “Architect output schema”The Architect produces a task DAG with expanded per-task fields.
| Property | Type | Requirement | Description |
|---|---|---|---|
tasks | array | Required | A list of task definitions that form the implementation plan. |
contract | object | Required | Structural definitions (e.g., GraphQL schemas, DB migrations) enforced across all tasks. |
validation | array | Required | Global validation rules for the implementation. |
cross_cutting_concerns | string[] | Required | Constraints that apply to multiple or all tasks in the plan. |
{ "tasks": [ { "id": "task_001", ... } ], "contract": { "graphql": "type Mutation { ... }" }, "validation": [ "Titles must be > 3 characters" ], "cross_cutting_concerns": [ "Use existing auth middleware" ]}Per-task object
Section titled “Per-task object”| Property | Type | Requirement | Description |
|---|---|---|---|
id | string | Required | Unique task identifier (e.g., task_001). |
title | string | Required | Concise name of the task. |
description | string | Required | Technical implementation details. |
acceptance_criteria | object[] | Required | Verifiable conditions for task completion. |
depends_on | string[] | Required | List of task IDs that must be completed first. |
files_touched | string[] | Required | Paths to files this task modifies or creates. |
{ "id": "task_001", "title": "Add Search Mutation", "description": "Implement searchBooks in src/backend/api/resolvers.py", "acceptance_criteria": [ { "criterion": "Mutation returns list of books", "verify_by": "test" } ], "depends_on": [], "files_touched": ["src/backend/api/resolvers.py"]}CSG JSON structure
Section titled “CSG JSON structure”The Codebase Semantic Graph (CSG) is stored at .speed/context/semantic-graph.json.
| Property | Type | Description |
|---|---|---|
nodes | object[] | Symbol definitions with schema metadata. |
edges | object[] | Relationships between symbols (calls, imports, etc.). |
clusters | object[] | Domain groupings derived from cohesion analysis. |
{ "id": "src/backend/models/user.py::User", "name": "User", "kind": "class", "file": "src/backend/models/user.py", "impact": { "blast_radius": 23, "stability": "bridge" }}Project Map schema
Section titled “Project Map schema”The Project Map acts as the canonical file registry at .speed/context/project-map.json.
| Property | Type | Description |
|---|---|---|
generated_at | string | ISO timestamp of the index build. |
git_head | string | The git commit hash at the time of the build. |
files | object[] | A list of file entries containing paths, languages, and line counts. |
{ "path": "src/backend/models/user.py", "language": "python", "lines": 340, "category": "source"}Spec-alignment schema
Section titled “Spec-alignment schema”Stored at .speed/features/{feature}/context/spec-alignment.json.
| Property | Type | Description |
|---|---|---|
claims | object[] | assertions extracted from the spec. |
status | enum | confirmed, missing, or divergent. |
evidence | string | Technical proof for the status. |
{ "source": { "spec_file": "specs/tech/auth.md", "line": 42 }, "claim_type": "entity_defined", "status": "confirmed", "evidence": "CSG: User class found at src/backend/models/user.py::User"}