Skip to main content
The Team page provides detailed profiles for each contributor, showcasing their expertise, coding patterns, and contribution history.

​
Overview

This page displays contributor profiles built from analyzing PR reviews and commit history. Each profile includes:
  • Contribution statistics
  • AI-generated profile summaries
  • Identified strengths
  • Common patterns and behaviors
  • Recent activity indicators
Team profiles are generated after at least 3 PR reviews from a contributor. The AI learns from code patterns, review feedback, and commit messages.

​
Page Header

The header shows:
  • Title: “Team”
  • Subtitle: “Developer profiles and contribution insights”
  • Repository Filter: Dropdown to select which repo’s contributors to display

​
Repository Selection

The dropdown lists all connected repositories: 
[Select Repository ▾]

owner/repo-1
owner/repo-2
organization/repo-3
Defaults to the first connected repository if not manually selected.

​
Summary Statistics

Four KPI cards at the top:

Contributors

Total number of developers who have opened PRs

Total PRs

Sum of all PRs across all contributors

Avg PR Count

Mean PRs per contributor (total PRs ÷ contributor count)

Active This Month

Contributors with at least one PR in the last 30 days

​
Contributor Cards

Each developer is displayed in a card with:

​
Header Section

  • Avatar Circle: Colored circle with username’s first letter (amber background)
  • Username: Bold @username label
  • PR Badge: Amber pill showing PR count (e.g., “28 PRs”)
  • Last PR: Monospace badge with PR number (e.g., “Last PR #456”)
Avatar colors:
  • Background: rgba(245, 192, 0, 0.1) (amber/10)
  • Border: rgba(245, 192, 0, 0.2) (amber/20)
  • Text: #F5C000 (amber)

​
Profile Summary

AI-generated paragraph describing the developer’s:
  • Primary areas of work
  • Coding style characteristics
  • Notable contributions
  • Technical focus areas
Example:
“Experienced full-stack developer focused on backend API design and database optimization. Consistently delivers well-tested code with clear documentation. Strong expertise in TypeScript, PostgreSQL, and REST API patterns.”
Profile summaries are generated using Claude 3.5 Sonnet by analyzing commit messages, PR descriptions, code changes, and review feedback.

​
Strengths Section

Green badges highlighting positive attributes:
  • Icon: ⭐ Star (success green)
  • Label: “STRENGTHS” in uppercase monospace
  • Up to 3 strength badges displayed
  • Text truncated at 35 characters + ellipsis
Example strengths: 
✓ Excellent test coverage
✓ Clean separation of concerns
✓ Comprehensive error handling
Badge styling:
  • Background: rgba(74, 219, 74, 0.1) (success/10)
  • Border: rgba(74, 219, 74, 0.2) (success/20)
  • Text: #4ADB4A (success green)

​
Patterns Section

Amber badges showing recurring behaviors or areas for improvement:
  • Icon: ⚠️ Alert triangle (amber)
  • Label: “PATTERNS” in uppercase monospace
  • Up to 2 pattern badges displayed
  • Text truncated at 35 characters + ellipsis
Example patterns: 
⚠ Tends to create large PRs (>500 LOC)
⚠ Occasional async/await misuse
Badge styling:
  • Background: rgba(245, 192, 0, 0.1) (amber/10)
  • Border: rgba(245, 192, 0, 0.2) (amber/20)
  • Text: #F5C000 (amber)
Patterns are observational, not prescriptive. They help teams identify coaching opportunities and knowledge gaps.

​
Card Layout

Cards are displayed in a responsive grid:
  • Desktop: 2 columns
  • Mobile: 1 column
  • Gap: 16px between cards

​
Hover Effects

  • Border color changes from border-surface-border to border-amber/20
  • Avatar background brightens to rgba(245, 192, 0, 0.2)
  • Smooth transition (all properties)

​
Empty State

If no contributor data exists: 
👥 No contributor data yet

Connect a repo and complete some PR reviews
to build contributor profiles.
  • Icon: Users icon (32px, gray)
  • Primary text: Bold, content-primary color
  • Secondary text: Smaller, content-secondary color
  • Centered layout with padding

​
Loading State

While fetching contributors:
  • 6 skeleton cards in a 2-column grid
  • Height: 192px per card (h-48 in Tailwind)
  • Background: #1a1a1a (surface-elevated)
  • Pulse animation

​
Data Source

​
API Endpoint

GET /api/analytics/contributors?repo={repo_full_name}
Response format: 
interface ContributorsResponse {
  repo: string;
  contributor_count: number;
  contributors: Contributor[];
}

interface Contributor {
  username: string;
  profile_summary: string | null;
  patterns: string[];
  strengths: string[];
  pr_count: number;
  commit_count: number;
  last_seen_pr: number | null;
}

​
Data Freshness

Contributor data updates:
  • When a new PR review completes
  • After repository rescan
  • When manually refreshing analytics
Run a repository rescan to rebuild contributor profiles with the latest data.

​
How Profiles Are Generated

Nectr builds contributor profiles through:

​
1. Data Collection

  • PR titles and descriptions
  • Commit messages and diffs
  • Code review comments
  • Files touched and languages used
  • Review verdicts and confidence scores

​
2. Pattern Analysis

The AI identifies:
  • Common coding mistakes
  • Preferred architectural patterns
  • Testing habits
  • Documentation practices
  • Communication style

​
3. Strength Detection

Positive patterns that indicate:
  • Technical expertise
  • Best practice adherence
  • Code quality focus
  • Collaboration skills

​
4. Summary Generation

Claude 3.5 Sonnet synthesizes all data into:
  • Concise profile paragraph
  • Top 3-5 strengths
  • Top 2-3 recurring patterns

​
Use Cases

Onboarding

New team members can quickly understand each developer’s expertise and working style.

Code Review Assignment

Assign reviewers based on strengths (e.g., security expert for auth changes).

Knowledge Gaps

Identify areas where the team lacks expertise and plan training.

1-on-1s

Use patterns as discussion points for coaching and growth conversations.

​
Privacy & Ethics

Contributor profiles are based on code patterns, not personal attributes. They should be used for technical coaching, not performance reviews.

​
Best Practices

  • Share profiles transparently with your team
  • Use patterns for constructive feedback
  • Celebrate strengths publicly
  • Discuss patterns privately in 1-on-1s
  • Update profiles regularly (monthly rescans)

​
What NOT to Do

  • Don’t use profiles for performance ratings
  • Don’t compare developers competitively
  • Don’t share patterns outside the team
  • Don’t treat AI insights as absolute truth

​
Technical Details

​
Profile Generation Prompt

Nectr uses a prompt template that includes: 
Analyze this developer's contribution history:

PRs: {pr_list}
Commits: {commit_list}
Reviews: {review_list}

Generate:
1. A 2-3 sentence profile summary
2. Top 3-5 technical strengths
3. Top 2-3 recurring patterns or areas for growth

Focus on code quality, architectural decisions, and collaboration.

​
Text Truncation

To prevent UI overflow:
  • Strengths truncated at 35 chars per badge
  • Patterns truncated at 35 chars per badge
  • Profile summary limited to 3 lines with line-clamp-3

​
Responsive Design

Breakpoints:
  • Mobile (< 640px): 1 column, full width
  • Tablet (640px - 1024px): 2 columns
  • Desktop (> 1024px): 2 columns with wider gutters

​
Future Enhancements

Upcoming features:
  • Search/Filter: Find developers by strength or pattern
  • Sortable Columns: Order by PR count, activity, etc.
  • Trend Charts: Visualize contribution growth over time
  • Expertise Tags: Categorize developers by technology stack
  • Team Composition: Aggregate view of team strengths/gaps
Request these features or suggest new ones on the Nectr GitHub repository.

Build docs developers (and LLMs) love

Get started for freeTalk to us