Parkour MCP

Parkour is a content exploration toolkit that helps LLMs surface high signal, unsummarized web content. It makes extensive use of clean APIs and Markdown conversion to enable targeted content extraction and knowledge synthesis. A rolling 2Q page cache keeps recently visited pages in memory so that follow-up requests (section extraction, BM25 search, slice retrieval, comparison pivots) are served quickly and without…

People who need web scraping tools inside Claude, Cursor, VS Code, or another MCP client. The project is written in Python.

VERIFIED ACTIVE

LAST COMMIT 2026-08-21 · ★ 4 · #62 OF 78 MAINTAINED WEB SCRAPING · VERIFIED 2026-08-25

MIT · Python servers · how we verify → /methodology

01 · Install Parkour MCP

before you install - you'll need

Set KAGI_API_KEY, S2_API_KEY, PLAYWRIGHT_BROWSER, MCP_AUTO_INSTALL_BROWSER, GITHUB_TOKEN, MCP_ALLOW_PRIVATE_IPS before connecting. S2_ACCEPT_TOS is optional or environment-specific per the README.

Claude Desktop

Settings → Extensions → Install, then select the https://github.com/blightbow/parkour-mcp/releases/download/v2.4.0/parkour-mcp.mcpb .mcpb bundle

02 · Evidence

Security posture

What to check before giving this server access to your agent - from the registry, GitHub, and our own probes. We don't score safety; we show what's verifiable.

runs as Claude Desktop extension (.mcpb bundle)

license MIT - declared in the repository

registry namespace io.github.blightbow is GitHub-verified and matches the repo owner

03 · What Parkour MCP can do

Prose above is summarized from the project's README and registry record - no invented capabilities.

Latest releases

v2.4.0 · 2026-08-21

Added · two settings now control JavaScript rendering - PLAYWRIGHT_BROWSER picks the engine (webkit, chromium, firefox, or auto) and MCP_AUTO_INSTALL_BROWSER=1 has the server download a missing browser on demand - and…

v2.3.0 · 2026-08-17

Changed · Reddit fetches no longer depend on curl-cffi, removing a dependency whose async streaming had an open unbounded-memory defect and whose browser-fingerprint updates are gated behind a commercial service.…

v2.2.2 · 2026-08-15

Fixed · BM25 search over a page or video transcript written in Japanese, Chinese, Korean, Thai or a neighbouring script silently returned no matches for any query, and now matches terms as they appear in the text, down…

04 · Who maintains Parkour MCP

parkour-mcp is maintained by blightbow. It's the only MCP server we track from this author; the repo dates to Jan 2026.

05 · Facts

repository: github.com/blightbow/parkour-mcp
category: web scraping - ranked #62 of 78 actively-maintained web scraping servers as of 2026-08-25.
release cadence: 10+ releases in the last 90 days (latest 2026-08-21)
registry: io.github.blightbow/parkour-mcp (active, first published 2026-04-10 · 21 versions)
packages: mcpb:https://github.com/blightbow/parkour-mcp/releases/download/v2.4.0/parkour-mcp.mcpb

06 · Parkour MCP FAQ

What is Parkour MCP?

Is Parkour MCP still maintained?

Yes - as of 2026-08-25, its last commit was 2026-08-21 and it shipped 10+ releases in the last 90 days. We re-verify nightly.

07 · Alternatives to Parkour MCP

Alternatives to Parkour MCP

Maintained web scraping servers if Parkour MCP isn't the fit.

Pairs well with

Servers that cover what Parkour MCP doesn't - only shown when the pairing reason fits the companion.

More web scraping MCP servers · Searchts · Amazon Scraper API · Jobseek · Crawleo MCP · Trentina

More Python MCP servers · Model Ledger · 13 FLOW · FreeCAD MCP · Agentguard47 · Gridbert · see all