Profiling Engineers via GitHub APIs | Generated by AI
Great question. Here’s a systematic framework for profiling an engineer through public APIs:
=== How to Know an Engineer Through Their GitHub ===
1. PROFILE API -- /users/{name}
curl -s 'https://api.github.com/users/{name}'
Reveals:
- bio: self-description (usually 1 line of identity)
- location: where they're based
- company: current employer
- blog: personal site (deep dive into this)
- created_at: how long they've been on GitHub
- public_repos / followers / following:
ratio tells you producer vs consumer
(many repos, few followers = builder, not promoter)
2. REPOS API -- /users/{name}/repos?sort=stars&per_page=100
curl -s 'https://api.github.com/users/{name}/repos?sort=stars&direction=desc'
Reveals:
- Languages: what they actually code in (not what they claim)
- Stars: what the community values from them
- Descriptions: how they think about their own work
- Topics: what domains they tag themselves in
- pushed_at: are they still active? or dormant since 2020?
- Forks vs originals: contributor or creator?
Patterns to look for:
- Dotfiles/config repos = tooling nerd
- Many tutorial/cheatsheet repos = learner/teacher
- One big project vs many small ones = deep vs broad
- Fork of popular project = contributing upstream
3. STARS API -- /users/{name}/starred?per_page=30
curl -s 'https://api.github.com/users/{name}/starred?per_page=30'
This is GOLD. Stars reveal:
- What they aspire to build (aspirational interests)
- What ecosystem they're investing in
- Their taste in tools and frameworks
- Emerging interests (recently starred)
Often more honest than repos -- repos show what they
finished, stars show what they care about.
4. EVENTS API -- /users/{name}/events?per_page=30
curl -s 'https://api.github.com/users/{name}/events?per_page=30'
Reveals:
- Recent activity: what they're working on RIGHT NOW
- Push events: which repos are actively developed
- Issue/PR activity: do they collaborate or solo?
- Watch events: what communities they follow
5. ORGS API -- /users/{name}/orgs
curl -s 'https://api.github.com/users/{name}/orgs'
Reveals:
- Company/team affiliations
- Open source communities they belong to
- Professional network
6. FOLLOWERS/FOLLOWING -- /users/{name}/followers
Reveals their professional network graph.
Who follows them = who values their work.
Who they follow = who they learn from.
7. PERSONAL SITE -- {blog} from profile
If they have blog/notes (like .github.io):
- Writing reveals depth of thinking
- Topics reveal what they care enough to write about
- Technical posts > "what I learned" posts for skill level
- Post frequency reveals consistency
8. LANGUAGE DISTRIBUTION
Don't just count repos -- count RECENT repos.
A dev with 20 Java repos but last 3 in Rust is
a Rust developer now, not a Java one.
Language timeline = career trajectory visible in data.
9. REPO NAMING & DESCRIPTIONS
- Professional descriptions = corporate background
- Emoji/quirky descriptions = startup/open-source culture
- Chinese descriptions = likely China-based, bilingual
- Academic terms = research background
- No description = just ships, doesn't market
10. CONTRIBUTION GRAPH (not in API, but visible)
github.com/{name} shows green squares:
- Consistent = disciplined, daily coder
- Bursty = project-driven, deadline-driven
- Gaps = job change, burnout, or side projects
- Weekend activity = passion project time
=== What Each Source Tells You ===
Source | Reveals
----------------+----------------------------------
Profile bio | Self-identity (1 line)
Repos | What they BUILT
Stars | What they ADMIRE / ASPIRE TO
Events | What they're doing NOW
Orgs | Who they WORK WITH
Followers | Who VALUES them
Blog | How they THINK
Languages | What they actually USE
Topics | What DOMAINS they care about
Forks | What they CONTRIBUTE TO
Stars (recent) | What they're LEARNING
Created_at | Seniority (years on platform)
=== Mutual Interest Detection ===
Compare two engineers:
1. Language overlap --> shared tech stack
2. Star overlap --> shared interests (strongest signal)
3. Topic overlap --> shared domains
4. Common orgs --> same communities
5. Common followers --> professional circle overlap
Star overlap is the most honest signal because:
- People star what genuinely interests them
- No social pressure (stars are private-ish)
- Shows aspiration, not just past work
Key insight: Stars > Repos for understanding someone’s interests. Repos show what they finished under constraints (work, time, opportunity). Stars show what they actually care about — their genuine curiosity. When I profiled nocmk2, the starred repos (hermes-agent, openclaw, langfuse, opencode) painted a much clearer picture of his AI-agent obsession than his 55 repos (many empty or forked).
Want me to turn this into a reusable script or skill that auto-profiles any GitHub user?