@drmaxbdc/productboard-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PRODUCTBOARD_AUTH_MODE | No | Set to 'oauth' to force OAuth even if PRODUCTBOARD_ACCESS_TOKEN is set; set to 'pat' to require PAT (good for CI). Default is auto-detect. | |
| PRODUCTBOARD_ACCESS_TOKEN | No | Personal Access Token for PAT authentication. If set, the MCP uses PAT and does not run the OAuth flow. | |
| PRODUCTBOARD_OAUTH_SCOPES | No | Space- or comma-separated scopes. Set this to bypass the chooser page. | |
| PRODUCTBOARD_OAUTH_CLIENT_ID | No | Your own OAuth app's client_id. Required for non-Dr.Max consumers until Productboard's dynamic registration endpoint works. | |
| PRODUCTBOARD_OAUTH_TOKEN_PATH | No | Override the tokens.json location (e.g. for Docker volumes). | |
| PRODUCTBOARD_OAUTH_CALLBACK_PORT | No | Override the callback port. Re-register the matching http://127.0.0.1:<port>/callback URI in your OAuth app. | 7779 |
| PRODUCTBOARD_OAUTH_CLIENT_SECRET | No | Required when using a Confidential Client (i.e. anything registered via PB's admin UI). Without it, token exchange returns HTTP 400. | |
| PRODUCTBOARD_OAUTH_REGISTRATION_PATH | No | Override the registration.json location. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_entity_configurationsA | Discover available entity types and their fields (field names, types, options, lifecycle operations). Call this before creating or updating entities to learn what fields are available. Optionally filter by entityType. |
| get_note_configurationsA | Discover available note types (simple, conversation, opportunity) and their fields. Call this before creating or updating notes. |
| list_entitiesA | List Productboard entities of a given type. Supports filtering by name, owner, status, archived, parent. Returns paginated results — use pageCursor from response to get next page. |
| get_entityA | Get a single Productboard entity by its UUID. Returns full entity data including fields and relationships. |
| create_entityB | Create a new Productboard entity. Call get_entity_configurations first to discover available fields for the entity type. The fields object should match the field schemas from configuration. |
| update_entityA | Update an existing Productboard entity. Use 'fields' for simple field replacement, or 'patch' for granular operations (set, addItems, removeItems, clear). These are mutually exclusive. Note: addItems/removeItems on 'teams' is emulated client-side (read-merge-set) due to a PB API bug. |
| delete_entityA | Permanently delete a Productboard entity. This cascades to all child entities and cannot be undone. |
| search_entitiesA | Search Productboard entities using POST with complex filters. Supports filtering by types, statuses, owners, parent, archived, and specific IDs. For name-based search use list_entities (its |
| get_entity_relationshipsA | Get all relationships for a Productboard entity (parent, children, links, blocking, etc.). Can optionally filter by relationship type or target. |
| create_entity_relationshipC | Create a relationship between two Productboard entities. |
| set_entity_parentA | Set or replace the parent of a Productboard entity. Uses PUT to replace any existing parent relationship. |
| delete_entity_relationshipC | Remove a relationship from a Productboard entity. Requires the relationship type and target entity ID. |
| list_notesA | List Productboard notes (also known as insights) with pagination and optional filters. Sorted by creation date (newest first). DEFAULT: returns processed + unprocessed notes that are NOT archived. To include archived notes set archived=true; to fetch only archived notes set archived=true and omit processed (v2 quirk: archived notes always report processed=false). |
| get_noteB | Get a single Productboard note by its UUID. |
| create_noteA | Create a new Productboard note (insight). Call get_note_configurations first to discover available fields and note types (simple, conversation). |
| update_noteB | Update an existing Productboard note. Use 'fields' for simple replacement or 'patch' for granular operations (set, clear, addItems, removeItems). |
| delete_noteB | Permanently delete a Productboard note. This cannot be undone. |
| get_note_relationshipsB | Get relationships for a Productboard note (linked customers, entities, etc.). |
| create_note_relationshipA | Link a Productboard note to a customer (user/company) or entity. For customer relationships, this replaces any existing customer link. |
| set_note_customerB | Set or replace the customer relationship on a note. Only accepts user or company targets. |
| delete_note_relationshipC | Remove a relationship from a Productboard note. |
| add_note_commentA | DEPRECATED — will stop working on 2026-07-08 when Productboard sunsets API V1. V2 has no equivalent note-comments endpoint as of 2026-05 (last confirmed via API changelog). No workaround available; the comment feature will be removed in v2.0.0 alongside the V1 client cleanup. Continue using only if comments are critical AND you have a contingency for sunset. |
| search_notesA | Search Productboard notes (insights). Routes to V2 POST /notes/search by default; falls back to V1 GET /notes only when term (fulltext) is set or allTags has 2+ values (V2 supports neither). The |
| list_all_notesA | Bulk-fetch Productboard notes via V2 API with auto-pagination (~100/page). Safety limit: 5000 notes max. DEFAULT: returns processed + unprocessed notes that are NOT archived. To include archived notes set archived=true. V2 response shape: each note has top-level {id, type, links{self,html}, fields{...}, relationships{...}, createdAt, updatedAt, metadata}. NOTE: v2 no longer returns followers[], embedded comments, totalResults, or features[].importance. Use links.html in place of v1 displayUrl. Linked features are now under /notes/{id}/relationships (not inline) — use get_note_relationships per note if you need them. |
| get_note_v1A | DEPRECATED: use get_note instead. Kept as an alias for backwards compatibility during the V1→V2 migration. Now calls v2 GET /notes/{id} (same as get_note). V1-only fields no longer available: followers[], embedded comments[], features[].importance. The v1 displayUrl is now exposed as links.html on the returned note. This tool will be removed in the next major release (v2.0.0-cleanup, after the V1 sunset on 2026-07-08). |
| resolve_noteA | Resolve a Productboard note from any identifier: UUID, numeric ID, web UI URL, or deep link. Returns the v2 note (with links.html — the web UI URL, equivalent to v1 displayUrl). For numeric IDs, scans up to 500 most-recent notes (~5 pages) matching against links.html. Includes archived notes in the scan. |
| list_membersB | List Productboard workspace members. Optionally filter by role. |
| get_memberB | Get a single Productboard workspace member by UUID. |
| get_member_activitiesB | Retrieve member activity metrics from Productboard. Returns daily activity data per member including feature/note/board creation counts. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 29 tools
The set covers distinct resource families (entities, notes, relationships, members), but note retrieval is fragmented across list_notes, list_all_notes, search_notes, get_note, get_note_v1, and resolve_note, which can cause misselection. The descriptions differentiate these paths with effort, but the overlap remains substantial enough to keep the score middling.
Tool names consistently follow a snake_case verb_noun pattern: list_, get_, create_, update_, delete_, search_, set_, add_, resolve_. The convention is predictable across all 29 tools, with only config/relationship nouns varying naturally by domain.
At 29 tools, the surface is heavy for the apparent scope and exceeds the 25-tool threshold for 'too many'. The count is further inflated by deprecated aliases and sunsetting tools such as get_note_v1 and add_note_comment, alongside overlapping note retrieval operations.
Core CRUD is covered for entities and notes, plus relationship lifecycle operations, member reads, activity metrics, and configuration discovery. Gaps remain: note comments have no V2 equivalent and are sunsetting, member write operations are absent, and the search_entities name filter is documented as non-functional upstream.