Webpage Integration: GitHub Activity Pipeline

[henryajisegiri]

Goal

Build the end-to-end pipeline to fetch store, and display a user's Boost GitHub activity on their profile page.

Figma Link

Important References
Spike Reference
Draft PR

---

Scope

GraphQL helper, data model and Celery sync task.

---

Background

The GitHub activity card on the user profile currently renders hardcoded placeholder content. Activity data should be sourced from GitHub using the existing app-level token (GITHUB_TOKEN) via GraphQL — not per-user OAuth — since Boost and cppalliance repositories are public and we already authenticate with a shared app token. Data is fetched for the boostorg org only.

Activity data is stored in the database and refreshed daily via a Celery task, so profile loads never call GitHub live.

---

Acceptance Criteria

GraphQL helper and activity fetcher

• A graphql(query, variables) method is added to GithubAPIClient, posting to the GitHub GraphQL endpoint using the app token (settings.GITHUB_TOKEN).
• A boost_activity(login) function runs the contributionsCollection query scoped to the boostorg org and returns the following:
  • Total commit count and number of repositories
  • Total repositories created
  • Total PRs opened and number of repositories
  • Total PR reviews and number of repositories
  • Featured PR (highest comment count): repository name, URL, and comment count
• The boostorg org node ID is stored as a settings constant to avoid resolving it on every call.
• A GraphQL or API error raises an exception rather than returning partial data silently.

Data model

• A GithubActivity model is added with a one-to-one relationship to User, storing the computed activity numbers, featured PR fields, and a last_synced timestamp.
• An upsert helper creates or updates the row for a given user.
• The model is visible in the Django admin for debugging.

Refresh task

• GitHub activity data is not refreshed by a scheduled nightly task for all users.
• When a user connects their GitHub account for the first time, the system performs an initial fetch of that user’s GitHub activity and stores it locally.
• When a profile page is opened, the system checks the stored GitHub activity record’s last-synced timestamp.
• If the stored data is not stale, the profile displays the existing stored GitHub activity without triggering a refresh.

Note: Stale = >24hours

• If the stored data is stale, the system triggers a refresh of that user’s GitHub activity and updates the stored record when the refresh completes.
• When a refresh is in progress, the UI shows a refreshing/loading indicator associated with the GitHub activity card.
• In the non-JS experience, the card shows the last updated state and communicates that a page refresh may be needed to see the newest data.
• When a user disconnects their GitHub account, the system deletes that user’s stored GitHub activity data.

Card display

• The GitHub activity card on the profile page reads from the stored GithubActivity row in the database. The page load does not call the GitHub API — all data is pre-fetched and stored by the background Celery task.
• If the user has no linked GitHub account, a connect prompt is shown on the card.
• Clicking "View on GitHub" CTA, redirects to the user's profile.

---

FE/BE Actions

Description	URL
Fetch stored GitHub activity for a user
Trigger single-user activity refresh on connect
Schedule daily activity sync task (Celery Beat)

---

Out of Scope

• Account connections card, connect/disconnect flows
• Per-user OAuth token storage
• Account connections card
• Email claim / commit author verification