FRESH
A shared URL-freshness oracle for AI agents. Returns REUSE, REFETCH, or UNKNOWN before an agent pays to retrieve a URL again.
README
FRESH
Know whether to fetch again.
FRESH is shared URL freshness intelligence for AI agents. Before re-fetching, re-scraping, re-rendering, or re-embedding a URL, ask whether the previously seen version is probably still fresh enough to reuse.
Production base URL: https://fresh-api-production-c783.up.railway.app
MCP endpoint: https://fresh-api-production-c783.up.railway.app/mcp
FRESH returns one of three decisions:
REUSE— cached knowledge is probably still fresh enoughREFETCH— the URL is likely stale enough to justify another retrievalUNKNOWN— evidence is insufficient; FRESH prefers uncertainty over false confidence
Why FRESH exists
A local cache knows when you last fetched something. It does not know whether the outside resource changed since then, nor what other callers recently observed. FRESH builds shared, privacy-safe URL change history from timestamps, ETags, Last-Modified values, and content hashes.
REST
POST /v1/check
{"url":"https://example.com/docs/api","lastSeenAt":"2026-08-13T12:00:00Z","toleranceSeconds":3600}
POST /v1/observe
{"url":"https://example.com/docs/api","observedAt":"2026-08-13T13:00:00Z","etag":"abc123","lastModified":"Wed, 13 Aug 2026 12:45:00 GMT","contentHash":"sha256:..."}
Raw page content is not required.
MCP
fresh_check— decide whether to retrieve a URL againfresh_observe— report privacy-safe freshness evidence after retrieval
Privacy
FRESH does not need raw page contents, cookies, target-site credentials, or customer payloads. URL keys are stored as one-way hashes with aggregate observation metadata.
Stranger verification
A core-tool invocation is evidence of use, not automatically proof of a genuine stranger. FRESH classifies candidate activity as KNOWN_VALIDATOR, LIKELY_VALIDATOR, CONTROLLED_TEST, UNKNOWN_MACHINE, or CREDIBLE_REAL_USE. Only CREDIBLE_REAL_USE advances stranger milestones.
Our acceptance/smoke traffic uses X-Tollbooth-Internal: 1 or X-Fresh-Internal: 1 so it cannot earn stranger credit.
Production acceptance gate
Every Railway deployment now performs live internal checks against the running service before /health can return 200. The gate exercises REST, UNKNOWN, REUSE, REFETCH, persistent observation reload, MCP initialize, MCP tool discovery, MCP fresh_check, and verifies that the controlled self-test does not increase the verified-stranger count.
Status
v0.1.2 experimental production infrastructure. Priorities: conservative decisions, low latency, sub-penny economics, privacy-safe shared learning, REST + MCP, durable observations, and auditable real-use analytics.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.