Package Registry API

Developer data from Python or the command line: npm search with weekly downloads and dependents, PyPI package metadata with dependencies and supported Python versions, crates.io search with download counts, Docker Hub images with pulls and stars, Hacker News stories and Stack Overflow questions, as JSON or CSV. The code is free and open source (MIT). Each of these registries and sites publishes its own API; this package puts all six behind one client and one QuanticData sign-in, returns the same flat rows and CSV export for every source, retries on rate limits and dropped connections, and runs long jobs in the background.

pip install git+https://github.com/quantumproxies/package-registry-api
package-registry login
package-registry npm "react state management" --max-results 20 -o npm.csv

login opens the QuanticData sign-in page in your browser. New accounts get $2 of free usage every month, no card required: create one.

What it scrapes

Command	Input	Returns	Per run
package-registry npm	search text	Version, license, weekly and monthly downloads, dependents, links, 21 fields per row (fields and price)	up to 100
package-registry pypi	package names	Version, license, Python support, dependencies, releases, links, 24 fields per row (fields and price)	up to 50
package-registry crates	crate name or keyword	Version, all-time and 90-day downloads, dates, docs and repository links, 17 fields per row (fields and price)	up to 100
package-registry docker	search text, or --images	Stars, pulls, official flag, description, pull command, 15 fields per row (fields and price)	up to 100
package-registry hackernews	search text (optional)	Title, link, points, comments, author, dates, 14 fields per row (fields and price)	up to 200
package-registry stackoverflow	keyword, or --tags	Score, answers, views, accepted answer, tags, license, 22 fields per row (fields and price)	up to 100

stackoverflow also reads Super User, Server Fault, Ask Ubuntu, MathOverflow, Unix, DBA, Security and Data Science with --site, and filters by tag with --tags "python;asyncio". hackernews sorts by relevance or date and keeps only Ask HN or Show HN posts with --type; crates sorts by relevance, downloads, recent downloads, recent updates or newest; docker reads exact images with --images redis bitnami/postgresql.

Sample output

A real row from package-registry npm "react state management" --max-results 3 (2026-10-23):

{
  "name": "unstated-next",
  "rank": 1,
  "license": "MIT",
  "npm_url": "https://www.npmjs.com/package/unstated-next",
  "version": "1.1.0",
  "bugs_url": null,
  "homepage": null,
  "insecure": false,
  "keywords": [],
  "publisher": "thejameskyle",
  "dependents": 257,
  "repository": null,
  "description": "200 bytes to never think about React state management libraries ever again",
  "maintainers": [
    "thejameskyle"
  ],
  "published_at": "2019-05-13T23:34:47.580Z",
  "search_score": 382.99,
  "quality_score": 1,
  "downloads_weekly": 85141,
  "popularity_score": 1,
  "downloads_monthly": 500976,
  "maintenance_score": 1
}

Python

from quanticdata_package_registry_api import PackageRegistry

registry = PackageRegistry()  # saved login, or QUANTICDATA_API_KEY

found = registry.npm("react state management", max_results=20)
for p in found:
    print(p["name"], p["version"], p["downloads_weekly"], p["dependents"])

meta = registry.pypi(["requests", "numpy", "fastapi"])
for p in meta:
    print(p["name"], p["version"], p["requires_python"], len(p["dependencies"]))

crates = registry.crates("http client", sort="downloads", max_results=10)
crates.to_csv("rust_http_clients.csv")

stories = registry.hackernews("sqlite", sort="date", max_results=30)
questions = registry.stackoverflow(tags="python;asyncio", sort="votes", max_results=30)
print(stories.rows[0]["title"], questions.rows[0]["title"])

Each call returns a Run: iterate it for the rows (plain dicts), take len(), or save it with to_csv(), to_json(), to_jsonl() or to_dataframe() (pandas). Errors are specific: InsufficientCredit, RateLimited, InvalidInput, JobFailed, AuthError.

Examples

• Dependency audit: Read a requirements.txt and get, for every package, the latest version on PyPI next to the one you pin, the Python versions it needs, the yanked flag and the known vulnerabilities PyPI lists. See examples/dependency_audit.py.
• npm landscape for a need: Search npm with a few phrases, merge the results into one list and sort it by weekly downloads, with the phrases that found each package. See examples/npm_landscape.py.
• What developers say about a tool: Put the Hacker News stories and the Stack Overflow questions about a library side by side, with score, replies, date and link. See examples/tech_mentions.py.

Command line

Results print as a table; -o file.csv, -o file.json or -o file.jsonl writes a file. List inputs take a b c, a,b,c, @file.txt (one per line) or - (stdin). --no-wait starts a long run in the background and prints its id; package-registry run <id> -o out.csv collects it later.

package-registry npm

Search the npm registry: name, latest version, description, license, weekly and monthly downloads, dependents, maintainers and links.

package-registry npm QUERY [options]

Option	Type	Default	Description
QUERY	string		Search text: a package name, words, keywords:web-scraping, author:NAME or a scope, e.g. "react state management".
--max-results	integer	20	How many packages to deliver at most (1–100). You pay only for delivered packages.

package-registry pypi

Python package metadata from PyPI — version, summary, license, Python support, links.

package-registry pypi PACKAGES [options]

Option	Type	Default	Description
PACKAGES	string[]		PyPI package names, e.g. requests, numpy.
--max-results	integer	10	How many packages to deliver at most (1–50). You pay only for delivered packages.

package-registry crates

Rust package search from the crates.io registry — downloads, versions, links.

package-registry crates QUERY [options]

Option	Type	Default	Description
QUERY	string		Crate name or keyword.
--sort	relevance / downloads / recent_downloads / recent_updates / new	relevance	Result order.
--max-results	integer	25	How many crates to deliver at most (1–100). You pay only for delivered crates.

package-registry docker

Docker image search and repository stats — stars, pulls, official flag.

package-registry docker [QUERY] [options]

Option	Type	Default	Description
QUERY	string		Image search text, e.g. postgres. Leave it out when you pass --images.
--images	string[]		Exact image references for detail lookups, e.g. redis or bitnami/postgresql.
--max-results	integer	25	How many images to deliver at most (1–100). You pay only for delivered images.

package-registry hackernews

Hacker News stories for a search query, or the current front page: title, link, points, comment count, author and dates.

package-registry hackernews [QUERY] [options]

Option	Type	Default	Description
QUERY	string		What to search, e.g. "web scraping". Omit it for the current front page.
--type	story / ask_hn / show_hn / front_page		Story type filter (default: story when searching, front_page otherwise).
--sort	relevance / date		relevance (default when searching) or date (newest first).
--max-results	integer	30	How many stories to deliver at most (1–200). You pay only for delivered stories.

package-registry stackoverflow

Questions from Stack Overflow and other Stack Exchange sites, by keyword or tag: score, answers, views, accepted answer, tags, license and, with --include-body, the question text.

package-registry stackoverflow [QUERY] [options]

Option	Type	Default	Description
QUERY	string		Keyword to search in titles and bodies, e.g. "web scraping".
--tags	string		Semicolon-separated tags, e.g. python;asyncio.
--site	stackoverflow / superuser / serverfault / askubuntu / mathoverflow / unix / dba / security / datascience	stackoverflow	Stack Exchange site (default stackoverflow).
--sort	votes / activity / creation / relevance		Sort order. Default: relevance when a query is given, votes for a tag-only listing.
--include-body	boolean	False	Also deliver the question body as plain text (same call, larger rows).
--max-results	integer	30	How many questions to deliver at most (1–100). You pay only for delivered questions.

Output fields

package-registry npm: 21 fields

Field	Type	Description
rank	integer	1-based position.
name	string	Package name.
version	string or null	Latest version.
description	string or null	Package description.
keywords	string[]	Declared keywords.
license	string or null	License (SPDX).
publisher	string or null	Last publisher.
homepage	string or null	Homepage URL.
repository	string or null	Repository URL.
npm_url	string	npmjs.com package URL.
published_at	string or null	Last publish (ISO 8601).
quality_score	number or null	Registry quality score 0–1; null while the registry sends a fixed placeholder (currently 1 for every package).
downloads_weekly	integer or null	Weekly downloads.
downloads_monthly	integer or null	Monthly downloads.
dependents	integer or null	Dependent packages.
maintainers	string[]	Maintainer usernames.
bugs_url	string or null	Issue tracker URL.
popularity_score	number or null	Registry popularity score 0–1; null while the registry sends a fixed placeholder (currently 1 for every package).
maintenance_score	number or null	Registry maintenance score 0–1; null while the registry sends a fixed placeholder (currently 1 for every package).
search_score	number or null	Relevance score of this result for the query.
insecure	boolean	Flagged insecure by the registry.

package-registry pypi: 24 fields

Field	Type	Description
rank	integer	1-based position.
name	string	Package name.
version	string or null	Latest version.
summary	string or null	Short summary.
author	string or null	Author name(s), never an e-mail address.
license	string or null	License.
requires_python	string or null	Required Python version spec.
homepage	string or null	Homepage URL.
repository	string or null	Source repository URL.
keywords	string[]	Declared keywords.
python_versions	string[]	Supported Python versions (classifiers).
release_count	integer or null	Number of releases.
url	string	PyPI project URL.
maintainer	string or null	Maintainer.
documentation	string or null	Documentation URL.
changelog	string or null	Changelog URL.
dependencies	string[]	Runtime dependencies (requires_dist without extras).
optional_dependencies_count	integer	Dependencies that belong to an extra.
extras	string[]	Extras the package provides.
development_status	string or null	Development status classifier.
released_at	string or null	Upload time of the current version.
first_release_at	string or null	Upload time of the first release.
yanked	boolean	Whether the current version is yanked.
known_vulnerabilities	integer or null	Known vulnerabilities PyPI lists for the current version.

package-registry crates: 17 fields

Field	Type	Description
rank	integer	1-based position.
name	string	Crate name.
description	string or null	Description.
version	string or null	Latest stable version.
downloads	integer or null	All-time downloads.
recent_downloads	integer or null	Downloads in the last 90 days.
created_at	string or null	First publish.
updated_at	string or null	Last update.
homepage	string or null	Homepage.
documentation	string or null	Docs URL.
repository	string or null	Repository URL.
exact_match	boolean	Query matched the name exactly.
url	string	crates.io page.
newest_version	string or null	Newest published version, pre-releases included.
versions_count	integer or null	Number of published versions.
yanked	boolean	Whether the crate's default version is yanked.
docs_rs_url	string	docs.rs page.

package-registry docker: 15 fields

Field	Type	Description
rank	integer	1-based position.
name	string	Image name (namespace/name; officials bare).
description	string or null	Short description.
stars	integer or null	Stars.
pulls	integer or null	Total pulls.
is_official	boolean	Docker official image.
is_automated	boolean	Automated build.
last_updated	string or null	Last push (detail mode only).
url	string	Docker Hub page.
namespace	string	Repository namespace (library for official images).
date_registered	string or null	Repository creation date (lookups).
categories	string[]	Docker Hub categories (lookups).
storage_size	integer or null	Bytes stored across all tags (lookups).
status	string or null	Repository status (lookups).
pull_command	string	docker pull command.

package-registry hackernews: 14 fields

Field	Type	Description
rank	integer	1-based position.
id	string	HN item id.
title	string	Story title.
url	string or null	Outbound link (null for text posts).
points	integer or null	Upvotes.
author	string or null	Submitter.
comments	integer or null	Comment count.
created_at	string or null	Submission time (ISO 8601).
text	string or null	Self-post text (Ask HN), tags stripped.
hn_url	string	Link to the HN discussion.
tags	string[]	Index tags (story, show_hn…).
domain	string or null	Host of the linked page (github.com); null for text posts.
updated_at	string or null	Last time the story changed in the index (ISO 8601).
type	string	story, ask_hn, show_hn, job or poll.

package-registry stackoverflow: 22 fields

Field	Type	Description
rank	integer	1-based position.
question_id	integer	Stack Exchange question id.
title	string	Question title (entities decoded).
score	integer or null	Net votes.
answer_count	integer or null	Number of answers.
view_count	integer or null	Views.
is_answered	boolean	Has at least one upvoted answer.
has_accepted_answer	boolean	An answer is accepted.
tags	string[]	Question tags.
author	string or null	Asker display name.
author_reputation	integer or null	Asker reputation.
created_at	string or null	Asked at (ISO 8601).
last_activity	string or null	Last activity (ISO 8601).
link	string or null	Question permalink.
accepted_answer_id	integer or null	Id of the accepted answer.
author_id	integer or null	Asker's user id on that site.
author_url	string or null	Asker's profile URL.
last_edit	string or null	Last edit time (ISO 8601).
closed_at	string or null	When the question was closed (ISO 8601).
closed_reason	string or null	Why it was closed (Duplicate, Needs more focus…).
license	string or null	Content license (CC BY-SA 4.0…).
body	string or null	Question body as plain text — only with include_body.

Sign-in and API keys

• package-registry login signs you in with OAuth 2.1 and PKCE in the browser. The token is stored in ~/.config/quanticdata/credentials.json (readable only by you) and refreshed automatically. One login serves every QuanticData tool on the machine.
• No browser on the machine (SSH, Docker, Colab): run package-registry login --no-browser, open the link anywhere, then paste back the address the browser ends on.
• Scripts and CI: set QUANTICDATA_API_KEY to a key from the dashboard, or store one with package-registry login --api-key.
• package-registry logout revokes the token. The connection is listed under API keys in the dashboard as "API · QuanticData CLI"; deleting that key disconnects every machine that uses it.

Pricing

The code is free and MIT-licensed. Runs are billed to your QuanticData account per delivered row, at the price shown on each collector page linked above; $2 of usage is free every month and failed runs are never billed. Plans with higher limits: pricing.

Other ways to call it

The same data over plain HTTP:

curl -X POST https://api.quanticdata.io/v1/scraper/collectors/npm_packages/run \
  -H "Authorization: Bearer $QUANTICDATA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"input": {"query": "react state management", "max_results": 3}}'

• API reference with Python, Node, PHP and curl examples.
• MCP server to give Claude, Cursor or any MCP client the same tools.
• n8n node for no-code workflows.

FAQ

Is this package registry API free?

The code is free and MIT-licensed. Runs go through your QuanticData account: $2 of usage is free every month with no card, then you pay per delivered row at the price shown on each collector page. Failed runs are never billed.

Do I need an API key for each source?

No. One QuanticData sign-in covers every command and the requests run on QuanticData's side. npm, PyPI, crates.io, Docker Hub, Hacker News and Stack Exchange publish their own APIs; what this package adds is one client for all of them, the same flat rows and CSV, JSON or JSONL export for every source, retries on rate limits and dropped connections, and background runs (--no-wait) you collect later with package-registry run <run_id>.

How many rows can one run return?

Up to 100 npm packages, 50 PyPI packages, 100 crates, 100 Docker images, 200 Hacker News stories or 100 Stack Overflow questions per run; ask for fewer with --max-results and you pay only for what is delivered.

Can I export to CSV or Excel?

Add -o results.csv (or .json, .jsonl). Lists such as keywords, maintainers, dependencies and tags are joined with '; ', so the CSV opens cleanly in Excel or Google Sheets.

Can I run it on a schedule?

Yes. Set QUANTICDATA_API_KEY in the environment and call the CLI or examples/dependency_audit.py from cron, GitHub Actions or any scheduler.

What about the usernames in the rows?

Rows carry the public usernames the sites show (npm publishers and maintainers, Hacker News submitters, Stack Overflow askers), and Stack Exchange content is licensed CC BY-SA, as each row's license field states. You are responsible for respecting each site's terms and the data-protection rules that apply to how you use the data.

Related scrapers

• GitHub Repository Search: GitHub repository search for Python and the command line.
• Public Data API: Public data API for Python and the command line.
• Web Scraper API: Web scraper API for Python and the command line.

License

MIT. Built and maintained by QuanticData.