Engineering, projects and design tools: delivery questions with evidence

Engineering, projects & design connectors

• Adobe Workfront

  Adobe Workfront reporting in plain language: projects, tasks, requests, logged hours, resource allocation and proof approvals, asked from Claude or Teams.

  Status: Available
• Aha!

  Aha! reporting without the pivot builder: ask about features, releases, initiatives, goals and idea votes in your Aha! workspaces from Claude or Slack.

  Status: Available
• Airtable

  Airtable MCP for analytics: ask about records, linked tables, single selects and dates across Airtable bases, under your own collaborator permissions.

  Status: Available
• Asana

  Asana MCP for governed analytics: ask about overdue tasks, project status updates, portfolio health and goal progress in Asana from Claude or Teams.

  Status: Available
• Azure DevOps

  Azure DevOps dashboard questions in plain language: cycle time, sprint carryover, pipeline pass rate and test results from Boards, Pipelines and Test Plans.

  Status: Available
• ClickUp

  ClickUp MCP for analytics on tasks, statuses, sprint points, time entries and custom fields across your ClickUp Spaces, Folders and Lists.

  Status: Available
• Figma

  Figma MCP for design operations: ask about component usage, detached instances, unresolved comments and file activity across Figma teams and projects.

  Status: Available
• GitHub

  GitHub MCP for engineering analytics: pull request cycle time, review wait, Actions failure rate and open security alerts across your GitHub organization.

  Status: Available
• GitHub Enterprise Server

  GitHub Enterprise Server API analytics for self-hosted instances: pull requests, reviews and Actions runs, reached on your network under each user's token.

  Status: Available
• GitLab

  GitLab MCP for DevSecOps analytics: merge request lead time, pipeline success rate, deployment frequency and vulnerability age across GitLab groups.

  Status: Available
• GitLab Self-Managed

  GitLab Self-Managed API analytics: merge requests, pipelines and deployments on your own host, reached through your network with tokens from your instance.

  Status: Available
• Jira Data Center

  Jira Data Center reporting from your own servers: issue age, sprint scope change, worklog hours and resolution time, asked in plain language with the JQL shown.

  Status: Available
• Linear

  Linear MCP for engineering analytics: cycle completion, issue lead time, triage backlog and project progress across Linear teams, with the GraphQL query shown.

  Status: Available
• Lucid

  Lucid API questions about Lucidchart and Lucidspark documents: who owns which diagrams, what changed recently and how folders and sharing are organized.

  Status: Available
• monday.com

  monday.com dashboard numbers asked in plain language: item counts by status, timeline slippage, workload by person and numbers column totals across boards.

  Status: Available
• Productboard

  Productboard AI questions about customer notes, feature demand by company, objectives and releases, asked in plain language from Claude, Slack or Teams.

  Status: Available
• Shortcut

  Shortcut reporting in plain language: story cycle time, iteration completion, epic progress and bug counts from the Shortcut project management tool.

  Status: Available
• Smartsheet

  Smartsheet dashboard questions in plain language: at-risk rows, % complete, schedule variance and budget columns across sheets in your workspaces.

  Status: Available
• Trello

  Trello AI questions over your boards: cards stuck in a list, overdue due dates, label counts and card throughput, computed from Trello actions.

  Status: Available

Questions that cross these systems

• What was median pull request cycle time by repository last month?
• Which projects have more than a quarter of tasks overdue?
• How many issues were reopened after release this quarter?

Questions and answers

Will engineering metrics single out individuals?

They do not need to. Questions will be answered at team, repository or project level, and an administrator can rule out per-person breakdowns.