v0.3.15  ·  20 tools live

MCP Server  ·  Open Standard  ·  MIT

Real-time web intelligence
for Claude.

Every result wrapped in a freshness envelope — so your agent always knows when it’s looking at data, not just what.

“I asked Claude to help me find a job. It gave me listings. I applied to three. Two didn’t exist anymore. One had closed two years ago. Claude had no idea.”

20
Tools
0
API Keys needed
5
Unique adapters
6h
Cron pipeline
1
Open Claude Desktop config
Mac: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
2
Add this entry
{ "mcpServers": { "freshcontext": { "command": "npx", "args": ["-y", "mcp-remote", "https://freshcontext-mcp.gimmanuel73.workers.dev/mcp"] } } }
3
Restart Claude Desktop
All 20 tools appear automatically. No Node, no Playwright, nothing to install.
✓Restart Claude Desktop after editing the config. Tools appear automatically.
Every result looks like this —
[FRESHCONTEXT] Source: https://github.com/anthropics/anthropic-sdk-python Published: 2026-04-11 Retrieved: 2026-04-13T07:42:00Z Confidence: high Score: 97/100 (current) --- Stars: 8.2k  | Language: Python | Last commit: 2 days ago Description: Anthropic’s official Python SDK for Claude API access... [/FRESHCONTEXT]

What’s included

20 tools. No API keys.

extract_landscape — extract_idea_landscape — extract_gov_landscape — extract_finance_landscape — extract_company_landscape
Five composite tools. One call fires 4–6 sources in parallel, assembles a unified timestamped report. Pass min_freshness_score: 70 to filter stale sections automatically.
extract_github
README, stars, forks, language, topics, last commit. Confidence high — sourced from GitHub API.
extract_hackernews
Top stories or Algolia search. Scores, comment counts, exact timestamps.
extract_arxiv
Official arXiv API. Papers, authors, submission dates, abstracts.
extract_reddit
Community sentiment from any subreddit via public JSON API.
extract_scholar
Google Scholar search results with years and author metadata.
extract_yc
YC company listings by keyword. Who’s funded in your space.
extract_producthunt
Recent launches by topic. Upvotes, comments, launch dates.
search_repos
GitHub repos ranked by stars. Activity signals — last push, open issues, language.
package_trends
npm and PyPI metadata. Version history, release cadence, last updated.
extract_finance
Live stock data via Yahoo Finance. Price, market cap, P/E, 52w range. Up to 5 tickers.
search_jobs
Real-time job listings from Remotive, RemoteOK, HN Hiring. Every listing freshness-badged.
extract_govcontracts
US federal contract awards from USASpending.gov. Company, amount, agency, period. No auth.
unique
extract_sec_filings
SEC 8-K material event disclosures. CEO changes, acquisitions, breaches — filed within 4 business days.
unique
extract_gdelt
Global news in 100+ languages via GDELT Project. Updated every 15 minutes.
unique
extract_gebiz
Singapore Government procurement tenders via data.gov.sg. Ministry of Finance open dataset.
unique
extract_changelog
Update history from any repo, npm package, or website. Auto-discovers /changelog paths.
unique

Real output

What Claude actually sees.

extract_github github.com/anthropics/anthropic-sdk-python
high97/100
[FRESHCONTEXT] Source: https://github.com/anthropics/anthropic-sdk-python Published: 2026-04-11 Retrieved: 2026-04-13T07:42:00Z Confidence: high Score: 97/100 (current) --- Description: The official Python library for the Anthropic API Stars: 8,241 | Forks: 721 | Language: Python Last commit: 2026-04-11 Topics: anthropic, claude, ai, llm, python-client --- README --- # Anthropic Python SDK The official Python library for the Anthropic API. ## Installation pip install anthropic ## Usage import anthropic client = anthropic.Anthropic() message = client.messages.create(...) [/FRESHCONTEXT]
Retrieved: 2026-04-13T07:42:00Z Decay rate: 1.0 (github)
extract_govcontracts query: "Palantir"
high89/100
[FRESHCONTEXT] Source: https://api.usaspending.gov/api/v2/search/spending_by_award/ Published: 2025-10-15 Retrieved: 2026-04-13T07:42:00Z Confidence: high Score: 89/100 (current) --- Federal contracts — Palantir [1] PALANTIR USG INC Amount: $227.4M Awarded: 2025-10-15 Period: 2025-10-15 → 2029-05-31 Agency: Department of Defense — CDAO / Army NAICS: 541511 — Custom Computer Programming Services Desc: MSS TASK ORDER — AI decision intelligence platform [2] PALANTIR USG INC Amount: $292.0M Awarded: 2024-06-18 Agency: Department of the Army Desc: MAVEN SMART SYSTEM — UI/UX Prototype [/FRESHCONTEXT]
Retrieved: 2026-04-13T07:42:00Z Source: USASpending.gov (official US Treasury)
search_jobs query: "TypeScript"  ·  remote_only: true
high99/100
[FRESHCONTEXT] Source: Remotive + RemoteOK + Arbeitnow + HN Hiring Retrieved: 2026-04-13T07:42:00Z Confidence: high Score: 99/100 (current) --- # Job Search: "TypeScript" · remote only Found: 24 listings — ● 8 fresh ● 11 recent ● 5 older [Remotive] Senior TypeScript Engineer — Vercel ● 1d ago — FRESH Location: Remote Worldwide | Type: full_time Apply: https://remotive.com/... [RemoteOK] TypeScript / React Engineer — Linear ● 3d ago — FRESH Salary: $140k–$190k | Location: Remote Apply: https://remoteok.com/... [HN Hiring] Staff Engineer (TypeScript) — Anthropic ● 5d ago — FRESH Posted in: Ask HN: Who is hiring? April 2026 [/FRESHCONTEXT]
Sorted freshest first — apply with confidence
extract_sec_filings query: "Palantir Technologies"
high74/100
[FRESHCONTEXT] Source: https://efts.sec.gov/LATEST/search-index Published: 2026-02-24 Retrieved: 2026-04-13T07:42:00Z Confidence: high Score: 74/100 (reliable) --- SEC 8-K Filings — Palantir Technologies 12 total filings found (showing 5) [1] PALANTIR TECHNOLOGIES INC Form: 8-K Filed: 2026-02-24 Period: 2026-02-21 View: https://www.sec.gov/cgi-bin/browse-edgar?... [2] PALANTIR TECHNOLOGIES INC Form: 8-K Filed: 2026-01-30 Period: 2026-01-28 Note: 8-K filings are legally mandated disclosures filed within 4 business days of material events. [/FRESHCONTEXT]
Source: SEC EDGAR (free, no auth, real-time) Not in any other MCP server

Under the hood

How freshness works.

1
Adapter retrieves
Each adapter calls a public API or scrapes a page. It returns raw content plus a content_date and a freshness_confidence level.
2
Stamp is applied
The FreshContext stamp wraps every result with a retrieval timestamp, source URL, confidence level, and a 0–100 freshness score based on domain-specific decay rates.
3
Agent can reason
Claude sees the envelope. It knows the data is 2 days old with high confidence — or 18 months old with low confidence. It can act accordingly, not blindly.
Freshness score formula —
freshness_score = max(0, 100 - (days_since_retrieved × decay_rate)) decay_rate by category: finance 5.0   (half-life ~10 days) jobs 3.0   (half-life ~17 days) news/HN 2.0   (half-life ~25 days) github 1.0   (half-life ~50 days) academic 0.3   (half-life ~167 days)