Built-in Tables

In addition to custom tables you create, every project includes read-only built-in tables in the abstra schema. These tables give you SQL access to workflow data like tasks, users, roles, and executions.

Read-only access

Built-in tables are accessible via SQL SELECT queries only. All writes go through the Abstra API and Python SDK.

Querying built-in tables

Use the abstra schema prefix when querying:

import abstra.tables as at

tasks = at.run_sql("SELECT * FROM abstra.tasks WHERE status = 'pending' LIMIT 10")
users = at.run_sql("SELECT * FROM abstra.users")

Available tables

abstra.tasks

Stores all workflow tasks and their lifecycle data.

Column	Type	Description
id	uuid	Task ID
created_at	timestamp	When the task was created
updated_at	timestamp	Last update
deleted_at	timestamp	Soft delete timestamp (null if active)
project_id	uuid	Project ID
type	varchar	Task type label used for routing
payload	jsonb	Task payload data
status	text	pending, locked, or completed
target_stage_id	varchar	Destination stage ID
created_by_stage_id	varchar	Stage that created this task
created_by_execution_id	uuid	Execution that created this task
completed_at	timestamp	When the task was completed
completed_by	uuid	Execution that completed this task
locked_at	timestamp	When the task was locked
locked_by	uuid	Execution that locked this task

abstra.executions

Records every stage execution in your workflow.

Column	Type	Description
id	uuid	Execution ID
created_at	timestamp	When the execution started
updated_at	timestamp	Last status change
status	text	created, running, success, failure, etc.
context	jsonb	Execution context (includes taskId if triggered by a task)
project_id	uuid	Project ID
build_id	uuid	Build that ran this execution
stage_id	text	Stage identifier
stage_run_id	uuid	Stage run ID (for forms threads)
worker_id	uuid	Worker that ran the execution
app_id	uuid	App ID (if applicable)
pid	text	Process ID

abstra.users

Stores access control users for your project.

Column	Type	Description
id	uuid	User ID
created_at	timestamp	When the user was added
updated_at	timestamp	Last update
deleted_at	timestamp	Soft delete timestamp
email	varchar	User email
project_id	uuid	Project ID

abstra.roles

Stores access control roles.

Column	Type	Description
id	uuid	Role ID
created_at	timestamp	When the role was created
updated_at	timestamp	Last update
deleted_at	timestamp	Soft delete timestamp
project_id	uuid	Project ID
name	varchar	Role name
description	varchar	Role description

abstra.users_roles

Junction table linking users to roles.

Column	Type	Description
id	uuid	Association ID
created_at	timestamp	When the association was created
deleted_at	timestamp	Soft delete timestamp
user_id	uuid	References abstra.users.id
role_id	uuid	References abstra.roles.id

abstra.signup_policy

Stores email pattern rules for user sign-up.

Column	Type	Description
id	uuid	Policy ID
project_id	uuid	Project ID
email_patterns	jsonb	Array of allowed email patterns (e.g. ["@company.com"])

Example queries

Pending tasks by stage

import abstra.tables as at

pending = at.run_sql("""
    SELECT target_stage_id, count(*) as pending_count
    FROM abstra.tasks
    WHERE status = 'pending' AND deleted_at IS NULL
    GROUP BY target_stage_id
""")

Recent failed executions

import abstra.tables as at

failures = at.run_sql("""
    SELECT id, stage_id, created_at, context
    FROM abstra.executions
    WHERE status = 'failure'
    ORDER BY created_at DESC
    LIMIT 20
""")

Users with their roles

import abstra.tables as at

users_with_roles = at.run_sql("""
    SELECT u.email, array_agg(r.name) as roles
    FROM abstra.users u
    LEFT JOIN abstra.users_roles ur ON u.id = ur.user_id AND ur.deleted_at IS NULL
    LEFT JOIN abstra.roles r ON ur.role_id = r.id AND r.deleted_at IS NULL
    WHERE u.deleted_at IS NULL
    GROUP BY u.email
""")