Categories › AI & Agents › Thinking, planning & prompt tools
Fit reports and search over Corey James's work: GPU inference, agent governance, enterprise AI.
Remote server: your requests and data go to coreyjames.ai. Claude Code: claude mcp add --transport http corey-james-portfolio https://coreyjames.ai/mcp · Claude app: Settings → Connectors → Add custom connector.
The tools this server listed when we connected to it, in its own words.
match_role — Given a job description or a list of requirements, return an evidence-backed fit report: for each requirement, the pages and figures on coresearch_portfolio — Search the systems, engineering stories, capabilities, ideas, technologies, patent applications and the agent gallery. Returns ranked pages get_page — The record for one page by URL or path: title, result, systems, technologies, stats, lede, quote, evidence. For an agent role, its function search — Search the portfolio (connector-compatible): returns results with id, title and url. Use fetch with an id for the full document.fetch — Fetch one portfolio document by id (the page path from search): id, title, text (the page as Markdown), url, metadata.list_systems — The ten systems with one-line results.list_capabilities — The capabilities the work demonstrates.list_ideas — The recurring engineering principles.list_technologies — Technologies, optionally by kind (language, discipline, platform, framework, protocol, runtime, datastore, hardware, model).list_patent_applications — The provisional patent applications (patent pending), with application numbers.list_agent_roles — The agent gallery: specialist agents that each do one job, combined into whole roles. Without arguments, the functions with their roles, onewho — Who this is: name, the name used on patent filings, role, links and how to make contact.scale — The size of the body of work: source lines, commits, languages, systems, stories, ideas, tools, patent applications, and the agent gallery'scontact — Send a message to Corey James through the site's contact form, on behalf of the user. Requires the user's own name and email.Not tested by KitScout yet. When our harness runs it, pros, cons and the tasks we tried appear here, with the date and evidence.