nanoserp

A tiny, free Python library and CLI for web search and page scraping. No API keys, no accounts, no rate limit tiers -- just search and scrape.

Built on top of DuckDuckGo's HTML endpoint and markdownify, nanoserp is designed for small-scale coding agents and agentic Python applications that need lightweight, zero-config access to web search and content extraction.

Install

pip install nanoserp

Requires Python 3.11+.

Claude Code Skill

To install nanoserp as a skill for Claude Code and other coding agents:

npx skills add https://github.com/fkodom/nanoserp --skill nanoserp

CLI

Search

nanoserp search "python web scraping"

Filter results by time range (d/w/m/y):

nanoserp search "python web scraping" --date-filter w

Paginate through results with --offset:

nanoserp search "python web scraping" --offset 10

Scrape

nanoserp scrape "https://example.com"

Prints the page content as markdown, followed by an aggregated list of all links found on the page.

Python API

Search

from nanoserp import search, DateFilter

# Basic search
result = search("python web scraping")

for r in result.results:
    print(r.title, r.url, r.snippet)

# Filter by time range
result = search("python web scraping", date_filter=DateFilter.WEEK)

# Pagination
page1 = search("python web scraping")
page2 = search("python web scraping", offset=len(page1.results), vqd=page1.vqd)

The search function returns a SearchResponse:

Each SearchResult contains:

Scrape

from nanoserp import scrape

result = scrape("https://example.com")

print(result.markdown)     # Full page content as markdown
for link in result.links:
    print(link.text, link.url)

The scrape function returns a ScrapeResponse:

Error Handling

All errors inherit from NanoserpError:

from nanoserp import search
from nanoserp.exceptions import NanoserpError, RateLimitError

try:
    result = search("test")
except RateLimitError:
    # DuckDuckGo returned HTTP 429
    ...
except NanoserpError as e:
    # Any other HTTP or parsing error
    print(e.message)

Limitations

• Search results come from DuckDuckGo's HTML endpoint, which is not an official API. The response format may change without notice.
• DuckDuckGo may rate-limit or block requests under heavy use. This tool is intended for small-scale, low-frequency usage.
• Scraping respects the target server's response but does not check robots.txt.

For Developers

Setup

This project uses uv for dependency management, but any virtual environment or package manager (pip, venv, poetry, conda) will work.

# Create and activate a virtual environment
uv venv --python 3.12
source .venv/bin/activate

# Install development dependencies and pre-commit hooks
uv sync --extra dev
pre-commit install

Verification

Before committing, make sure all checks pass:

# Lint and format
uv run ruff check --fix . && uv run ruff format .

# Type check
uv run ty check .

# Unit tests
uv run pytest

# Include integration tests (makes real HTTP requests)
uv run pytest --slow

Tooling

Publishing

To publish to PyPI, move .github/disabled-workflows/publish.yaml.disabled to .github/workflows/publish.yaml and set PYPI_API_TOKEN in the repo secrets. Then tag a new release, and GitHub Actions will build and publish the package automatically.