GitHub search (beta)

POST /v1/github/search

Search Ploid's index of GitHub accounts. Every criterion is a filter, combined with AND; give at least one.

curl https://api.ploid.com/v1/github/search \
  -H "Authorization: Bearer $PLOID_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "repo": "vercel/next.js",
    "min_commits": 100,
    "active_since": "2026-01-01",
    "sort": "commits",
    "limit": 10,
    "include_profiles": true
  }'

Criteria

Results are ranked by relevance when query or company is set, otherwise by merged pull requests, then commits. Merged pull requests need someone else to accept the work, so mirrors and automated accounts rarely rank high on them. sort ranks by one activity measure instead: recent_activity is pushes in the last 12 months, and last_active is the latest public activity.

Activity data

Activity comes from public GitHub events in GH Archive, 2011 to the latest index build. Repositories count under their current name, so work on zeit/next.js counts for vercel/next.js. Private work is not visible.

• Commits and pull requests stop at October 2025, when GitHub stopped publishing them in public events. Commits are those the account pushed, including merges, so maintainers who merge others' work count high. Commits pushed to forks are left out, since syncing a fork pushes its upstream's history, and a repository counts at most 20 commits per push, so an imported history does not read as years of work.
• Pushes run to the latest build and are the measure of recent work. From 2026, public events carry far fewer pull requests and reviews, so those counts understate recent activity.
• Automation is a best guess: GitHub's bot accounts, logins such as release-bot or nightly-builds, and accounts pushing more than 10,000 times a year. Some team and CI accounts with ordinary logins still get through.
• Languages are the main language of each repository the account contributed to, weighted by its commits and pull requests there.

Response

{
    "data": {
        "results": [
            {
                "login": "ada",
                "github_id": 1234,
                "type": "User",
                "company": "Stripe",
                "hireable": false,
                "has_email": true,
                "site_admin": false,
                "github_star": false,
                "repos": ["vercel/next.js", "stripe/stripe-node"],
                "repo_count": 2,
                "profile_url": "https://github.com/ada",
                "activity": {
                    "commits": 1840,
                    "pushes": 1210,
                    "pushes_last_12m": 96,
                    "pull_requests": 412,
                    "merged_pull_requests": 350,
                    "reviews": 288,
                    "active_months": 71,
                    "first_active_at": "2016-03-02",
                    "last_active_at": "2026-10-09",
                    "languages": ["TypeScript", "Go"],
                    "top_repos": [
                        {
                            "repo": "vercel/next.js",
                            "commits": 920,
                            "pushes": 610,
                            "pull_requests": 240,
                            "merged_pull_requests": 210,
                            "last_active_at": "2026-10-09"
                        }
                    ]
                },
                "profile": {
                    "username": "ada",
                    "name": "Ada Lovelace",
                    "bio": "...",
                    "location": "San Francisco",
                    "followers": 420,
                    "topLanguages": ["TypeScript", "Go"],
                    "topRepos": []
                }
            }
        ],
        "returned": 1,
        "total": { "value": 1, "relation": "eq" },
        "offset": 0,
        "limit": 10,
        "next_offset": null
    },
    "meta": { "request_id": "req_...", "credits_charged": 2 }
}

repos lists up to 100 repositories. activity is null for an account with no public activity on record, and top_repos lists its ten busiest repositories. profile appears only with include_profiles and is null when GitHub returns no profile for the account. Emails and phone numbers are never returned; find them with POST /v1/enrich.

Page with next_offset, up to 1,000 results deep. The last page before that limit shrinks to fit, so next_offset can always be sent back with the same limit. total counts up to 10,000 matches exactly; above that, relation is gte.

Pricing

A search costs 1 ACU, and each live profile found costs 1 ACU more, charged together when the search completes. Rejected and failed searches are not charged.

GitHub to LinkedIn

POST /v1/github/people-search

Run the same GitHub search, read each account's live profile, and tie it to a LinkedIn profile. Then filter on the LinkedIn side, for example to keep only senior engineers in the Bay Area.

curl https://api.ploid.com/v1/github/people-search \
  -H "Authorization: Bearer $PLOID_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "repo_owner": "vercel",
    "type": "User",
    "linkedin": {
      "location": { "any_of": ["San Francisco", "New York"] },
      "seniority": "senior"
    }
  }'

The request takes every GitHub criterion above (at least one is required), plus:

Each result has github (the account with its live profile), linkedin, and match. A match is one of:

Two different people tied at the best evidence are ambiguous, so neither is linked, and linkedin and match are null. A listed LinkedIn profile that Ploid has not indexed yet comes back with indexed: false and empty fields, and cannot pass linkedin filters. Matching uses Ploid's people index and the GitHub profile only; it runs no web search.

Pages advance over GitHub accounts: scanned counts the accounts a page examined, and next_offset moves past them, so a filtered page can return fewer people than it scanned. A search costs 1 ACU plus 1 ACU per returned person, charged together.

Errors

See Errors & rate limits.