Both files had their Last reviewed markers refreshed today without the content gaps being closed, so the freshness signal is currently green on two docs that are missing real material. azure-devops.md — `ado.reconcile_interval` is undocumented. It is real, user-tunable config (internal/ado/reconcile.go, default 10) setting how many syncs pass between automatic reconciliation scans, and the doc already documents the `--reconcile` flag that forces one. Adds a Reconciliation section, and notes that `ado.syncs_since_reconcile` is beads' own counter rather than a setting. otel-data-model.md — claims a complete schema but lists 7 of 111 storage spans, so the issueops role surface is invisible to anyone instrumenting from it. Rather than 100+ prose sections that would drift again, state the rule the code follows — span name is "storage." + operation name, which is also the db.operation attribute (InstrumentedStorage.op) — and inventory both families: 72 direct storage.<Method> spans and 39 storage.<Role>.<Method> spans. Also records why the families do not double-count: the role decorators wrap the UNINSTRUMENTED store (IssueLifecycle calls Unwrap() before wrapping), so one logical operation emits one span on either path and the counters stay comparable. No behaviour change; documentation catching up to the code. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
11 KiB
title, description
| title | description |
|---|---|
| Azure DevOps (ADO) Integration Configuration | Configuration reference for bd ado sync, which bidirectionally syncs beads issues with Azure DevOps work items |
Last reviewed: 2026-08-07
Freshness source: cmd/bd/ado*.go and internal/ado/.
This guide covers all configuration options for the bd ado sync command, which synchronizes beads issues with Azure DevOps work items.
Proxied-server mode: bd ado sync, bd ado status, and bd ado projects
are not supported when bd is connected to a proxied server (error:
ado sync is not supported in proxied-server mode); run them from a
workspace with direct database access.
Quick Start
# Set required config
bd config set ado.pat "your-personal-access-token"
bd config set ado.org "your-organization"
bd config set ado.project "your-project"
# Or use environment variables
export AZURE_DEVOPS_PAT="your-personal-access-token"
export AZURE_DEVOPS_ORG="your-organization"
export AZURE_DEVOPS_PROJECT="your-project"
# Sync (bidirectional)
bd ado sync
# Pull only (import from ADO)
bd ado sync --pull-only
# Push only (export to ADO)
bd ado sync --push-only
# Preview without making changes
bd ado sync --dry-run
Connection Configuration
| Config Key | Env Variable | Required | Description |
|---|---|---|---|
ado.pat |
AZURE_DEVOPS_PAT |
Yes | Personal Access Token |
ado.org |
AZURE_DEVOPS_ORG |
Conditional¹ | Organization name (e.g., myorg) |
ado.url |
AZURE_DEVOPS_URL |
Conditional¹ | Custom base URL (on-prem ADO Server) |
ado.project |
AZURE_DEVOPS_PROJECT |
Conditional² | Single project name |
ado.projects |
AZURE_DEVOPS_PROJECTS |
Conditional² | Comma-separated project names |
¹ Either ado.org or ado.url must be set. Use ado.url for on-premises Azure DevOps Server.
² At least one project must be configured via ado.project or ado.projects.
Config vs env var precedence: Config keys (set via bd config set) take priority over environment variables.
ado.pat is a secret key stored in config.yaml (per-repo or user-global),
never the shared database, so the PAT cannot leak via dolt push.
On-Premises ADO Server
For Azure DevOps Server (on-prem), use ado.url instead of ado.org:
bd config set ado.url "https://tfs.company.com/DefaultCollection"
bd config set ado.project "MyProject"
Multi-Project Sync
Sync across multiple projects in a single command:
bd config set ado.projects "ProjectA,ProjectB,ProjectC"
The first project is used as the primary for URL construction. WIQL queries use TeamProject IN (...) for multi-project support.
Filter Configuration
Filters control which ADO work items are included in sync operations.
| Config Key | CLI Flag | Description | Example |
|---|---|---|---|
ado.filter.area_path |
--area-path |
Area path (uses UNDER) | Project\Team |
ado.filter.iteration_path |
--iteration-path |
Sprint/iteration path | Project\Sprint 1 |
ado.filter.types |
--types |
Work item types (comma-separated) | Bug,Task,User Story |
ado.filter.states |
--states |
ADO states (comma-separated) | New,Active,Resolved |
CLI flags override config values for that sync run.
WIQL query example (generated from filters):
SELECT [System.Id] FROM WorkItems WHERE
[System.TeamProject] = 'MyProject'
AND [System.IsDeleted] = false
AND [System.AreaPath] UNDER 'Project\Team'
AND [System.WorkItemType] IN ('Bug', 'Task')
AND [System.State] IN ('New', 'Active')
ORDER BY [System.ChangedDate] ASC
Default Mappings
Priority Mapping
Priority mapping is bidirectional but lossy for P3/P4:
| Beads Priority | ADO Priority | Direction | Notes |
|---|---|---|---|
| 0 (Critical) | 1 | ↔ | |
| 1 (High) | 2 | ↔ | |
| 2 (Medium) | 3 | ↔ | Default for unknown values |
| 3 (Low) | 4 | → | |
| 4 (Backlog) | 4 | → | Lossy: becomes P3 on pull |
Note: Beads P3 and P4 both map to ADO priority 4. On a fresh pull into an empty database, ADO 4 maps back to beads P3. The original priority is not preserved across a full round-trip for P4 issues.
For Bug-type work items, ADO also requires a Severity field:
| Beads Priority | ADO Severity |
|---|---|
| 0 | 1 - Critical |
| 1 | 2 - High |
| 2 | 3 - Medium |
| 3, 4 | 4 - Low |
Status Mapping
| Beads Status | Default ADO State | Config Key |
|---|---|---|
open |
New |
ado.state_map.open |
in_progress |
Active |
ado.state_map.in_progress |
blocked |
Active + beads:blocked tag |
ado.state_map.blocked |
deferred |
Removed |
ado.state_map.deferred |
closed |
Closed |
ado.state_map.closed |
Blocked status: ADO has no native blocked state. beads maps blocked to Active and adds a beads:blocked tag. On pull, Active + beads:blocked tag restores StatusBlocked.
Override defaults for your process template:
# Example: Scrum template
bd config set ado.state_map.open "To Do"
bd config set ado.state_map.in_progress "In Progress"
bd config set ado.state_map.closed "Done"
Type Mapping
| Beads Type | Default ADO Type | Config Key |
|---|---|---|
bug |
Bug |
ado.type_map.bug |
feature |
User Story |
ado.type_map.feature |
task |
Task |
ado.type_map.task |
epic |
Epic |
ado.type_map.epic |
chore |
Task |
ado.type_map.chore |
Reverse mapping (ADO → beads) also recognizes:
Product Backlog Item→feature(Scrum template)Issue→task
Override for your process template:
# Example: Scrum template
bd config set ado.type_map.feature "Product Backlog Item"
Process Template Configuration
ADO supports multiple process templates with different work item types and state transitions. The defaults assume the Agile template. Override mappings for other templates.
Agile (Default)
No configuration needed. Default mappings work out of the box.
State transitions:
Bug: New → Active → Resolved → Closed
Task: New → Active → Closed
User Story: New → Active → Resolved → Closed
Epic: New → Active → Resolved → Closed
Scrum
bd config set ado.type_map.feature "Product Backlog Item"
bd config set ado.state_map.open "New"
bd config set ado.state_map.in_progress "Committed"
bd config set ado.state_map.closed "Done"
State transitions:
Product Backlog Item: New → Approved → Committed → Done
Task: To Do → In Progress → Done
Bug: New → Approved → Committed → Done
CMMI
bd config set ado.type_map.feature "Requirement"
bd config set ado.state_map.open "Proposed"
bd config set ado.state_map.in_progress "Active"
bd config set ado.state_map.closed "Closed"
State transitions:
Requirement: Proposed → Active → Resolved → Closed
Task: Proposed → Active → Closed
Bug: Proposed → Active → Resolved → Closed
State Transition Handling
When creating a work item in a non-initial state (e.g., pushing a closed issue), beads:
- Creates the item in the initial state (e.g.,
New) - Transitions through intermediate states to reach the target
- Example: Creating a closed Bug →
New → Active → Resolved → Closed
If a direct transition fails (ADO returns 400), beads automatically walks the known transition path for the work item type and process template.
Sync Options
Direction
| Flag | Description |
|---|---|
| (none) | Bidirectional: pull then push |
--pull-only |
Import from ADO only |
--push-only |
Export to ADO only |
Conflict Resolution
When the same issue has been modified both locally and in ADO:
| Flag | Description |
|---|---|
--prefer-newer |
Most recently updated version wins (default) |
--prefer-local |
Local beads version always wins |
--prefer-ado |
ADO version always wins |
Additional Flags
| Flag | Description |
|---|---|
--dry-run |
Preview sync without making changes |
--no-create |
Only update existing items, never create new ones |
--bootstrap-match |
Enable heuristic title matching for first sync |
--reconcile |
Force reconciliation scan for deleted items (see Reconciliation below) |
--issues |
Sync specific issues by bead ID or ADO work item ID |
--parent |
Push only a bead and its descendants (push mode only; mutually exclusive with --issues) |
--project |
Project name(s) for this run, repeatable — overrides ado.project/ado.projects |
--states |
Filter by work item states (comma-separated) |
--types |
Filter by work item types (comma-separated) |
Reconciliation
A reconciliation scan re-checks work items beads already tracks, so items deleted
in ADO (404) or no longer readable (403) are detected instead of lingering. It
does not run on every sync — that would cost an API call per tracked item — so it
runs periodically, and --reconcile forces it immediately.
| Config Key | Default | Description |
|---|---|---|
ado.reconcile_interval |
10 |
Number of syncs between automatic reconciliation scans |
ado.syncs_since_reconcile also appears in config: it is the counter beads keeps
to know when the next scan is due, not a setting to edit.
PAT Permissions
The Personal Access Token needs these scopes:
| Scope | Access | Required For |
|---|---|---|
| Work Items | Read & Write | Creating and updating work items |
Generate a PAT at: https://dev.azure.com/{org}/_usersettings/tokens
Metadata Preserved
beads stores ADO-specific metadata for round-trip fidelity:
| Metadata Key | Description |
|---|---|
ado.rev |
ADO revision number |
ado.area_path |
Area path |
ado.iteration_path |
Iteration/sprint path |
ado.story_points |
Story points estimate |
ado.remaining_work |
Remaining work hours |
ado.severity |
Bug severity value |
Description Conversion
- Push (beads → ADO): Markdown converted to HTML
- Pull (ADO → beads): HTML converted to Markdown
Tags and Labels
- ADO tags are semicolon-separated; beads labels use arrays
- User labels round-trip through ADO tags
- Internal
beads:*tags (e.g.,beads:blocked) are filtered on pull — they don't appear as user labels
API Limits
| Limit | Value |
|---|---|
| Max batch size | 200 work items per GET request |
| Max response size | 50 MB |
| Request timeout | 30 seconds |
| Max retries | 3 (GET and WIQL only) |
| Retry backoff | Exponential with jitter, respects Retry-After header |
Troubleshooting
Common Errors
ado.pat not configured: set via 'bd config set ado.pat <token>' or AZURE_DEVOPS_PAT env var
bd config set ado.pat "your-pat-here"
# or
export AZURE_DEVOPS_PAT="your-pat-here"
ado.org not configured: set via 'bd config set ado.org <org>' or AZURE_DEVOPS_ORG env var
bd config set ado.org "your-org"
# or for on-prem:
bd config set ado.url "https://tfs.company.com/DefaultCollection"
State transition errors (400 Bad Request)
This usually means the process template doesn't support a direct state change. Check your ado.state_map.* config matches your actual process template.
Type not found errors
Verify your ado.type_map.* config matches the work item types available in your project. Use --types filter to restrict which types are synced.
Debugging
# Preview what would happen
bd ado sync --dry-run
# Check current config
bd config get ado.pat
bd config get ado.org
bd config get ado.project