Summary
Wekan's board composite publication exposes full integration objects, including webhook URLs and authentication tokens, to any client subscribed to a board.
Because board publications are available to all board members and even unauthenticated clients on public boards, any user who can access a board may retrieve webhook credentials and invoke connected external services without authorization.
Affected context
The vulnerability occurs in Wekan deployments that:
- Run versions 8.31.0 through 8.33
- Use board integrations such as webhooks or external service connectors
- Rely on Meteor's DDP publications to deliver board data to clients
In this context, the board composite publication sends integration records to all subscribers without field-level filtering, so secrets like webhook URLs and tokens are exposed to any client that can subscribe to the board publication, including low-privilege members and unauthenticated users on public boards.
Exploit preconditions
- Attack surface: Wekan's Meteor/DDP API and client subscriptions for board publications
- Required access:
- For private boards: any board membership (e.g., read-only, comment-only)
- For public boards: ability to connect as an unauthenticated DDP client
- Trigger point: Subscribing to the vulnerable board composite publication for a target board
- Why exploitation is possible: The publication returns entire integration documents with no field-level filtering, so sensitive fields (webhook URLs, authentication tokens, shared secrets) are included in the data sent to the client, regardless of the client's role or authentication state.
Why this happens
This issue stems from insecure data exposure in a server-to-client publication layer. The Meteor composite publication responsible for serving board data includes related integration documents but does not perform field projection or redaction for sensitive properties.
As a result, it violates the principle of least privilege by providing all integration fields to any subscriber, mapping directly to CWE-200: Exposure of Sensitive Information to an Unauthorized Actor. The access control dimension is reflected by CWE-862: Missing Authorization, because the system fails to enforce that only privileged roles should see secret tokens used to authenticate with external webhooks and services.
Potential impact
- Attackers may obtain webhook URLs and authentication tokens for integrations tied to a board.
- Unauthorized parties could invoke external webhooks as if they were the Wekan system.
- External services integrated with Wekan could process attacker-controlled actions or data.
- Activity logs or audit trails in external systems may falsely attribute actions to Wekan.
- In some configurations, attackers could pivot into broader workflows or automation chains triggered by compromised webhooks.
General mitigation guidance
Effective mitigation focuses on tightening data exposure and enforcing authorization boundaries around secrets:
- Apply field-level filtering on server publications, returning only non-sensitive fields to clients and excluding tokens, shared secrets, and internal identifiers.
- Introduce separate data paths for secrets, accessible only to backend components or highly privileged administrative APIs, not to general-purpose client subscriptions.
- Enforce role-based access control at the publication layer so low-privilege users and anonymous clients never receive integration credentials.
- Use scoped, revocable tokens for webhooks and external integrations so that any exposed credential can be rotated quickly with minimal blast radius.
- Adopt a secure-by-default schema where secret fields are explicitly marked as confidential and require opt-in exposure only under controlled conditions.
How teams detect issues like this
To detect similar issues, organizations can:
- Perform code reviews focused on publication/subscription logic, checking that:
- All data returned by publications uses explicit field projections.
- Sensitive fields (tokens, secrets, credentials, API keys) are never included in generic publications.
- Use static analysis or linting rules that flag:
- Publication handlers returning complete documents without projections.
- Known secret-field names being passed directly from database queries to publications.
- Leverage application-level logging and monitoring to:
- Track which roles and clients are receiving sensitive fields.
- Detect anomalous access patterns to integration data or unusual DDP subscriptions.
- Incorporate secret scanning on outbound payloads in pre-production test environments to detect accidental exposure of tokens in client-visible responses.

