Accesses stored web content for analysis without live fetching, reducing latency and cost while leveraging cached data for insights.
The agent queries the AfterImage MCP server to retrieve cached versions of web pages. It checks the llms.txt endpoint for available resources and uses the cache to fetch content. Before acting, the agent should verify that the required pages are within the 10-day eviction window and that the user has sufficient credits for any live fetches if the cache is stale.
- Category: Research and knowledge
- Effect: Read-only
- MCP servers: AfterImage MCP (https://koskamo.com/explore/afterimage-mcp)
- Source: AfterImage — docs (https://afterimage.page/docs)
- Source quote: "Agents start at https://afterimage.page/llms.txt . AfterImage is a cache, not an archive."
- Playbook page: https://koskamo.com/explore/use-cases/retrieve-cached-web-pages-for-research
Prompt: Retrieve the cached content for the following URLs: [URL1], [URL2]. If the cache is expired, perform a live fetch. Summarize the key findings from the retrieved pages.