Database Schema

Foundation uses Forge SQL (MySQL 5.7 compatible) for all relational data and Forge KVS (key-value store) for session data, counters, and user preferences.

Current schema version: 32 (defined in foundation/src/db/migrations.ts)

Entity Relationships
Tables
Migration System
Storage Split: SQL vs KVS

Entity Relationships

lenses (1) ──────── (*) hierarchy_nodes
   │                        │
   │                        └── jira_issue_id -> issue_cache.jira_issue_id
   │
   ├──── (*) generators
   ├──── (*) views
   ├──── (*) lens_permissions
   ├──── (*) ai_insights
   ├──── (*) baselines ──── (*) baseline_entries
   ├──── (*) leveling_jobs
   ├──── (*) dependency_lag_overrides
   ├──── (*) lens_snapshots
   ├──── (*) automation_rules
   └──── (*) alerts

resources (1) ──── (*) resource_schedules
   │              │
   │              └── schedule_id -> work_schedules.id
   │
   ├──── (*) team_memberships ──── team_id -> teams.id
   ├──── (*) resource_absences
   ├──── (*) resource_calendar_assignments ──── calendar_id -> holiday_calendars.id
   └──── (*) resource_skills ──── skill_id -> skills.id

holiday_calendars (1) ──── (*) holiday_dates

Tables

`lenses`

Top-level entity. Each lens is a curated view of Jira issues organized as a hierarchy tree.

Column	Type	Default	Description
`id`	`CHAR(36)` PK		UUID v4
`numeric_id`	`INT` UNIQUE		Auto-assigned sequential ID for friendly URLs
`name`	`VARCHAR(255)` NOT NULL		Display name
`description`	`TEXT`		Optional description
`mode`	`VARCHAR(16)`	`'power'`	Mode: `power`, `deleting`
`owner_account_id`	`VARCHAR(128)` NOT NULL		Jira account ID of the owner
`item_limit`	`INT`	`1000`	Max issues allowed in this lens
`archived`	`INTEGER`	`0`	0=active, 1=archived
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`
`updated_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

`hierarchy_nodes`

Tree structure for a lens. Each node is either a Jira issue, a flex item (grouping node), a milestone, or a sync agent group.

Column	Type	Default	Description
`id`	`CHAR(36)` PK		UUID v4
`lens_id`	`CHAR(36)` NOT NULL		Parent lens
`parent_id`	`CHAR(36)`		Parent node (null = root level)
`position`	`INT` NOT NULL		Sort order within parent (gap=100 strategy)
`node_type`	`VARCHAR(32)` NOT NULL		`issue`, `flex`, `generator_group`, `milestone`
`jira_issue_id`	`VARCHAR(64)`		Jira issue ID (for `issue` nodes)
`jira_issue_key`	`VARCHAR(32)`		Jira issue key (e.g., `PROJ-123`)
`flex_name`	`VARCHAR(255)`		Display name (for `flex`/`milestone` nodes)
`flex_description`	`TEXT`		Description (for `flex` nodes)
`flex_icon`	`VARCHAR(128)`		Icon identifier (for `flex` nodes)
`hidden`	`BOOLEAN`	`FALSE`	Hidden by filter sync agents
`generator_id`	`CHAR(36)`		Links to the sync agent that created this node
`milestone_date`	`DATE`		Target date (for `milestone` nodes)
`scheduling_mode`	`VARCHAR(16)`	`'auto_bottom_up'`	Scheduling mode for auto-scheduling
`leveling_delay_days`	`INT`	`0`	Leveling delay in days
`depth`	`INT` NOT NULL	`0`	Tree depth (0 = root)
`created_by`	`VARCHAR(128)`		Jira account ID of creator
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`
`updated_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_lens_parent (lens_id, parent_id, position) — tree traversal
idx_lens_issue (lens_id, jira_issue_id) UNIQUE — prevents duplicate issues per lens
idx_jira_key (jira_issue_key) — issue key lookups
idx_hierarchy_generator_id (generator_id) — sync agent lookups

`generators`

Sync agent configurations. Each generator defines a JQL query or rule that populates a lens.

Column	Type	Default	Description
`id`	`CHAR(36)` PK		UUID v4
`lens_id`	`CHAR(36)` NOT NULL		Parent lens
`parent_node_id`	`CHAR(36)`		Target parent node for generated issues
`generator_type`	`VARCHAR(32)` NOT NULL		Type: `jql_insert`, `child_extend`, `hierarchy_builder`, `jpd_links`, `auto_project`, etc.
`position`	`INT` NOT NULL		Execution order
`config`	`JSON` NOT NULL		Type-specific config (JQL query, project key, etc.)
`created_by`	`VARCHAR(128)`		Creator’s account ID
`last_executed_at`	`TIMESTAMP`		Last execution timestamp
`last_execution_status`	`VARCHAR(16)`		`success`, `error`, `partial`
`last_execution_message`	`TEXT`		Human-readable status message
`last_known_total`	`INT`	`NULL`	JQL result count (for freshness checks)
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_lens (lens_id)

`issue_cache`

Local cache of Jira issue data. Kept fresh by product event triggers. This is the primary read source for the frontend.

Column	Type	Default	Description
`jira_issue_id`	`VARCHAR(64)` PK		Jira issue ID
`jira_issue_key`	`VARCHAR(32)` NOT NULL		Issue key (e.g., `PROJ-123`)
`project_id`	`VARCHAR(64)` NOT NULL		Jira project ID
`project_key`	`VARCHAR(16)` NOT NULL		Project key (e.g., `PROJ`)
`issue_type_id`	`VARCHAR(32)`		Issue type ID
`issue_type_name`	`VARCHAR(128)`		Issue type name (Story, Bug, etc.)
`summary`	`VARCHAR(512)` NOT NULL		Issue summary/title
`status_id`	`VARCHAR(32)`		Status ID
`status_name`	`VARCHAR(128)`		Status display name
`status_category`	`VARCHAR(32)`		`new`, `indeterminate`, `done`
`assignee_account_id`	`VARCHAR(128)`		Assignee’s Jira account ID
`assignee_display_name`	`VARCHAR(255)`		Assignee name
`assignee_avatar_url`	`VARCHAR(512)`		Assignee avatar URL
`reporter_account_id`	`VARCHAR(128)`		Reporter’s Jira account ID
`reporter_display_name`	`VARCHAR(255)`		Reporter name
`reporter_avatar_url`	`VARCHAR(512)`		Reporter avatar URL
`resolution_id`	`VARCHAR(32)`		Resolution ID
`resolution_name`	`VARCHAR(128)`		Resolution name (Done, Won’t Do, etc.)
`resolution_date`	`TIMESTAMP`		When the issue was resolved
`priority_id`	`VARCHAR(16)`		Priority ID
`priority_name`	`VARCHAR(64)`		Priority name (Highest, High, etc.)
`sprint_id`	`VARCHAR(32)`		Current sprint ID
`sprint_name`	`VARCHAR(128)`		Current sprint name
`sprint_state`	`VARCHAR(32)`		`active`, `closed`, `future`
`parent_issue_id`	`VARCHAR(64)`		Jira parent issue ID (Epic)
`story_points`	`DECIMAL(10,2)`		Story points / estimation
`labels`	`JSON`		Array of label strings
`components`	`JSON`		Array of component objects
`fix_versions`	`JSON`		Array of fix version objects
`affects_versions`	`JSON`		Array of affects version objects
`due_date`	`DATE`		Due date
`start_date`	`DATE`		Start date (promoted from `customfield_10015` in v15)
`created_at`	`TIMESTAMP`		Issue creation date in Jira
`updated_at`	`TIMESTAMP`		Last update date in Jira
`cached_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`	When this cache entry was last refreshed
`custom_fields`	`JSON`		All non-standard custom fields as JSON blob
`links_json`	`TEXT`		Issue links as JSON array
`description_text`	`TEXT`		Plain text from ADF description (max 500 chars)
`environment_text`	`TEXT`		Plain text from ADF environment (max 500 chars)
`original_estimate_seconds`	`INT`		Original time estimate
`remaining_estimate_seconds`	`INT`		Remaining time estimate
`time_spent_seconds`	`INT`		Time spent
`hierarchy_level`	`INT`		Jira hierarchy level (-1=Subtask, 0=Story, 1=Epic)
`last_accessed_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`	For auto-expiry (v32)

Indexes:

idx_project (project_key) — project-scoped queries
idx_parent (parent_issue_id) — parent issue lookups
idx_key (jira_issue_key) — key-based lookups

`views`

Saved view configurations for a lens (column order, formatting, Gantt settings).

Column	Type	Default	Description
`id`	`CHAR(36)` PK		UUID v4
`lens_id`	`CHAR(36)` NOT NULL		Parent lens
`name`	`VARCHAR(255)` NOT NULL		View display name
`is_default`	`BOOLEAN`	`FALSE`	Whether this is the default view
`visibility`	`VARCHAR(16)`	`'public'`	`public` or `private`
`owner_account_id`	`VARCHAR(128)` NOT NULL		Creator’s account ID
`columns`	`JSON` NOT NULL		Column configuration array
`settings_json`	`TEXT`		Additional settings (density, Gantt config, formatting, etc.)
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`
`updated_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_lens (lens_id)

`lens_permissions`

Access control list for lenses.

Column	Type	Default	Description
`id`	`CHAR(36)` PK		UUID v4
`lens_id`	`CHAR(36)` NOT NULL		Target lens
`grantee_type`	`VARCHAR(16)` NOT NULL		`user`, `group`, `role`, `everyone`
`grantee_id`	`VARCHAR(128)`		Account ID, group name, or role name (null for `everyone`)
`permission_level`	`VARCHAR(32)` NOT NULL		`view`, `edit`, `control`
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_lens (lens_id)
idx_unique_grant (lens_id, grantee_type, grantee_id) UNIQUE — note: NULL != NULL in SQL, so everyone grants need separate dedup (see v31 migration)

`ai_insights`

AI-generated portfolio insights.

Column	Type	Description
`id`	`CHAR(36)` PK	UUID v4
`lens_id`	`CHAR(36)` NOT NULL	Target lens
`insight_type`	`VARCHAR(32)` NOT NULL	Type of insight
`severity`	`VARCHAR(16)` NOT NULL	`info`, `warning`, `critical`
`summary`	`TEXT` NOT NULL	Human-readable summary
`affected_nodes`	`TEXT`	JSON array of affected node IDs
`suggested_action`	`TEXT`	JSON suggested action object
`created_at`	`BIGINT` NOT NULL	Unix timestamp (ms)
`resolved_at`	`BIGINT`	Unix timestamp when resolved

Indexes:

idx_ai_insights_lens (lens_id, created_at DESC)

`resources`

Resource profiles (linked to Jira users or placeholder roles).

Column	Type	Default	Description
`id`	`VARCHAR(36)` PK		UUID v4
`jira_account_id`	`VARCHAR(128)`		Jira account ID (null for placeholders)
`display_name`	`VARCHAR(255)`		Display name
`email`	`VARCHAR(255)`		Email address
`avatar_url`	`VARCHAR(512)`		Avatar URL
`hours_per_day`	`DECIMAL(4,2)`	`8.00`	Available hours per day
`efficiency_multiplier`	`DECIMAL(3,2)`	`1.00`	Productivity multiplier
`is_placeholder`	`BOOLEAN`	`FALSE`	Whether this is a placeholder resource
`placeholder_role`	`VARCHAR(255)`		Role name (for placeholders)
`is_active`	`BOOLEAN`	`TRUE`	Active status
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`
`updated_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_resources_account (jira_account_id) UNIQUE

`work_schedules`

Work schedule templates (hours per weekday).

Column	Type	Default
`id`	`VARCHAR(36)` PK
`name`	`VARCHAR(255)` NOT NULL
`hours_monday` - `hours_sunday`	`DECIMAL(4,2)`	8.00 (M-F), 0.00 (S-S)
`is_default`	`BOOLEAN`	`FALSE`

`resource_schedules`

Links resources to work schedules for specific date ranges.

Column	Type
`id`	`VARCHAR(36)` PK
`resource_id`	`VARCHAR(36)` NOT NULL
`schedule_id`	`VARCHAR(36)` NOT NULL
`start_date`	`DATE` NOT NULL
`end_date`	`DATE`

`teams`

Team definitions.

Column	Type	Default
`id`	`VARCHAR(36)` PK
`name`	`VARCHAR(255)` NOT NULL
`description`	`VARCHAR(512)`
`color`	`VARCHAR(7)`	`'#4C9AFF'`
`is_global`	`BOOLEAN`	`TRUE`
`lens_id`	`VARCHAR(36)`	(null for global teams)
`default_hourly_rate`	`DECIMAL(10,2)`
`created_at` / `updated_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

`team_memberships`

Team member assignments with availability.

Column	Type	Default
`id`	`VARCHAR(36)` PK
`team_id`	`VARCHAR(36)` NOT NULL
`resource_id`	`VARCHAR(36)` NOT NULL
`availability_pct`	`DECIMAL(5,2)`	`100.00`
`start_date`	`DATE` NOT NULL
`end_date`	`DATE`

`holiday_calendars`

Named holiday calendars (optionally region-specific).

Column	Type
`id`	`VARCHAR(36)` PK
`name`	`VARCHAR(255)` NOT NULL
`region`	`VARCHAR(64)`
`is_default`	`BOOLEAN` (default `FALSE`)
`created_at`	`TIMESTAMP`

`holiday_dates`

Individual holiday entries in a calendar.

Column	Type
`id`	`VARCHAR(36)` PK
`calendar_id`	`VARCHAR(36)` NOT NULL
`holiday_date`	`DATE` NOT NULL
`name`	`VARCHAR(255)`

`resource_absences`

Resource absence records (vacation, sick leave, etc.).

Column	Type	Default
`id`	`VARCHAR(36)` PK
`resource_id`	`VARCHAR(36)` NOT NULL
`start_date`	`DATE` NOT NULL
`end_date`	`DATE` NOT NULL
`is_half_day`	`BOOLEAN`	`FALSE`
`absence_type`	`VARCHAR(64)`	`'vacation'`
`notes`	`VARCHAR(512)`
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

`resource_calendar_assignments`

Links resources to holiday calendars.

Column	Type
`id`	`VARCHAR(36)` PK
`resource_id`	`VARCHAR(36)` NOT NULL
`calendar_id`	`VARCHAR(36)` NOT NULL
`start_date`	`DATE`
`end_date`	`DATE`

`baselines`

Point-in-time snapshots for schedule comparison.

Column	Type
`id`	`VARCHAR(36)` PK
`lens_id`	`VARCHAR(36)` NOT NULL
`name`	`VARCHAR(255)` NOT NULL
`created_by`	`VARCHAR(128)`
`created_at`	`TIMESTAMP`

`baseline_entries`

Individual node snapshots within a baseline.

Column	Type
`id`	`VARCHAR(36)` PK
`baseline_id`	`VARCHAR(36)` NOT NULL
`node_id`	`VARCHAR(36)` NOT NULL
`jira_issue_key`	`VARCHAR(32)`
`start_date`	`DATE`
`end_date`	`DATE`

`skills`

Skill taxonomy (global, not per-lens).

Column	Type	Default
`id`	`VARCHAR(36)` PK
`name`	`VARCHAR(255)` NOT NULL
`category`	`VARCHAR(128)`
`color`	`VARCHAR(7)`	`'#8777D9'`
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_skills_name (name) UNIQUE

`resource_skills`

Links resources to skills with proficiency levels.

Column	Type	Default
`id`	`VARCHAR(36)` PK
`resource_id`	`VARCHAR(36)` NOT NULL
`skill_id`	`VARCHAR(36)` NOT NULL
`proficiency_level`	`SMALLINT`	`3` (scale 1-5)
`start_date`	`DATE`
`end_date`	`DATE`

`leveling_jobs`

Async resource leveling job tracking.

Column	Type	Default
`id`	`VARCHAR(36)` PK
`lens_id`	`VARCHAR(36)` NOT NULL
`status`	`VARCHAR(16)`	`'pending'`
`started_at`	`TIMESTAMP`
`completed_at`	`TIMESTAMP`
`tasks_leveled`	`INT`	`0`
`error_message`	`TEXT`
`result_json`	`TEXT`
`created_by`	`VARCHAR(128)`
`created_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

`dependency_lag_overrides`

Per-lens dependency metadata (lag days, dependency type).

Column	Type	Default
`id`	`VARCHAR(36)` PK
`lens_id`	`VARCHAR(36)` NOT NULL
`from_issue_key`	`VARCHAR(32)` NOT NULL
`to_issue_key`	`VARCHAR(32)` NOT NULL
`jira_link_id`	`VARCHAR(64)`	Jira issue link ID
`dependency_type`	`VARCHAR(2)` NOT NULL	`'FS'` (FS, SS, FF, SF)
`lag_days`	`INT` NOT NULL	`0`
`updated_by`	`VARCHAR(128)`
`updated_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

Indexes:

idx_dep_lag_unique (lens_id, from_issue_key, to_issue_key) UNIQUE
idx_dep_lag_link_id (jira_link_id)

`monitoring_metrics`

Resolver performance metrics.

Column	Type	Default
`id`	`INT` AUTO_INCREMENT PK
`metric_name`	`VARCHAR(128)` NOT NULL
`metric_value`	`FLOAT` NOT NULL
`metric_type`	`VARCHAR(32)`	`'gauge'`
`tags_json`	`TEXT`
`recorded_at`	`TIMESTAMP`	`CURRENT_TIMESTAMP`

`lens_snapshots`

Daily health snapshots for trend tracking.

Column	Type	Default
`id`	`VARCHAR(64)` PK
`lens_id`	`VARCHAR(64)` NOT NULL
`snapshot_date`	`VARCHAR(10)` NOT NULL	YYYY-MM-DD
`total_items`	`INTEGER`	`0`
`completed_items`	`INTEGER`	`0`
`overdue_items`	`INTEGER`	`0`
`unassigned_items`	`INTEGER`	`0`
`total_points`	`REAL`	`0`
`completed_points`	`REAL`	`0`
`health_score`	`INTEGER`	`0`
`status_breakdown`	`TEXT`	JSON
`assignee_breakdown`	`TEXT`	JSON
`created_at`	`BIGINT` NOT NULL	Unix timestamp

`automation_rules`

User-configured automation rules.

Column	Type	Default
`id`	`VARCHAR(64)` PK
`lens_id`	`VARCHAR(64)` NOT NULL
`rule_type`	`VARCHAR(64)` NOT NULL
`config`	`TEXT` NOT NULL	JSON
`created_by`	`VARCHAR(128)` NOT NULL
`enabled`	`INTEGER`	`1`
`last_triggered_at`	`INTEGER`	Unix timestamp
`trigger_count`	`INTEGER`	`0`
`created_at`	`INTEGER` NOT NULL	Unix timestamp

`alerts`

System-generated and user-visible alerts.

Column	Type	Default
`id`	`VARCHAR(64)` PK
`lens_id`	`VARCHAR(64)` NOT NULL
`account_id`	`VARCHAR(128)` NOT NULL	Target user
`alert_type`	`VARCHAR(64)` NOT NULL
`severity`	`VARCHAR(16)` NOT NULL
`title`	`VARCHAR(512)` NOT NULL
`detail`	`TEXT`
`related_issue_key`	`VARCHAR(32)`
`suggested_action`	`TEXT`
`acknowledged`	`INTEGER`	`0`
`created_at`	`INTEGER` NOT NULL	Unix timestamp
`acknowledged_at`	`INTEGER`

Migration System

How It Works

ensureSchema() runs on every Forge cold-start (called by withSchema guard or inline in resolvers)
It reads the current schema version from KVS key foundation:schema_version
If the version is current, it returns immediately
If no version exists (fresh install), it runs createFreshSchema() which executes all CREATE TABLE IF NOT EXISTS statements
If an older version exists, it calls runMigrations(currentVersion, CURRENT_SCHEMA_VERSION) which sequentially executes each migration’s up() function
A KVS lease (lease:foundation:schema-bootstrap) prevents concurrent bootstrap from multiple cold-starts

Adding a New Migration

Checklist:

Bump CURRENT_SCHEMA_VERSION in foundation/src/db/migrations.ts

Add ONE migration entry to the migrations array:

{
  version: 33,
  description: 'Add new_column to some_table',
  up: async () => {
    await safeDDL('ALTER TABLE some_table ADD COLUMN new_column VARCHAR(128)');
  },
},

Add the column to the fresh-install CREATE TABLE in foundation/src/db/schema.ts
Use safeDDL() for ALTER TABLE / RENAME TABLE / CHANGE COLUMN (idempotent on re-run)
Use sql.executeDDL() for CREATE TABLE IF NOT EXISTS (inherently idempotent)
Run migration tests: cd foundation && npx jest --testPathPattern='schema|migration' --no-coverage

`safeDDL()` vs `sql.executeDDL()`

Function	Use For	Behavior on Re-run
`safeDDL()`	ALTER TABLE, RENAME TABLE, CHANGE COLUMN	Swallows errors (idempotent)
`sql.executeDDL()`	CREATE TABLE IF NOT EXISTS, CREATE INDEX IF NOT EXISTS	Inherently idempotent

Notable Migrations

Version	Description
v14	Renamed `structures` -> `lenses`, `structure_id` -> `lens_id` across 9 tables
v15	Promoted `start_date` from `custom_fields` JSON to dedicated column
v23	Added `numeric_id` to lenses with sequential backfill
v27	Deduplicated same-lens issue nodes + added UNIQUE constraint
v30	Collapsed `edit_generators` permission level into `edit`
v31	Deduplicated `everyone` grants (NULL != NULL workaround)
v32	Added `last_accessed_at` for auto-expiry

Storage Split: SQL vs KVS

Forge SQL

All structured, relational data that needs querying, indexing, or joins:

Lenses, hierarchy nodes, generators, views, permissions
Issue cache, dependency overrides, baselines
Resources, teams, schedules, skills, absences
AI insights, alerts, automation rules, monitoring metrics, snapshots

Forge KVS

Ephemeral, per-user, or counter data that doesn’t need relational queries:

Key Pattern	Data	TTL
`foundation:schema_version`	Current schema version (integer)	Permanent
`user_prefs:${accountId}`	User preferences (JSON)	Permanent
`announcements_dismissed:${accountId}`	Dismissed announcement IDs (array)	Permanent
`support_chat:${accountId}`	Support chat messages	Permanent
`rate_limit:hourly:*`	Hourly point usage shards	1 hour (logical)
`tier:budget_override`	Admin-set tier budget	Permanent
`active-gen-job:${lensId}`	Active sync job marker	15 min (logical)
`job:${jobId}`	Async job status	Permanent
`cacheRefreshJob:${jobId}`	Cache refresh job status	Permanent
`admin_refresh_last:global`	Last cache refresh timestamp	Permanent
`admin:css_override`	Custom CSS (string)	Permanent
`admin:reset_in_progress`	Reset crash recovery flag	Cleared on completion
`project_components:${projectKey}`	Cached project components	1 hour TTL
`project_versions:${projectKey}`	Cached project versions	1 hour TTL
`last-issue-event-ts:${projectKey}`	Last issue event timestamp	Permanent
`browse_check:${project}:${user}`	BROWSE permission cache	30 min TTL
`feature_flags:*`	Feature flag values	Permanent
`lease:*`	Distributed lock leases	TTL-based

Database Schema

Table of Contents

Entity Relationships

Tables

lenses

hierarchy_nodes

generators

issue_cache

views

lens_permissions

ai_insights

resources

work_schedules

resource_schedules

teams

team_memberships

holiday_calendars

holiday_dates

resource_absences

resource_calendar_assignments

baselines

baseline_entries

skills

resource_skills

leveling_jobs

dependency_lag_overrides

monitoring_metrics

lens_snapshots

automation_rules

alerts

Migration System

How It Works

Adding a New Migration

safeDDL() vs sql.executeDDL()

Notable Migrations

Storage Split: SQL vs KVS

Forge SQL

Forge KVS

`lenses`

`hierarchy_nodes`

`generators`

`issue_cache`

`views`

`lens_permissions`

`ai_insights`

`resources`

`work_schedules`

`resource_schedules`

`teams`

`team_memberships`

`holiday_calendars`

`holiday_dates`

`resource_absences`

`resource_calendar_assignments`

`baselines`

`baseline_entries`

`skills`

`resource_skills`

`leveling_jobs`

`dependency_lag_overrides`

`monitoring_metrics`

`lens_snapshots`

`automation_rules`

`alerts`

`safeDDL()` vs `sql.executeDDL()`