Jev AtlasField notesSearch⌘K

Machine interface · v0.1

Give your agent the field guide.

Jev Atlas gives coding agents independent research, evidence status, project examples, architecture patterns, and build hypotheses—without making them scrape the visual website or load the entire corpus.

81normalized records
9tracked claims
31opportunities
5read-only tools

01 Connect

One endpoint, bounded context.

The MCP is public and read-only. It does not call Jev, accept TypeSafe credentials, modify repositories, or execute external instructions.

Remote MCP configuration
{
  "mcpServers": {
    "jev-atlas": {
      "url": "https://YOUR-DOMAIN/mcp"
    }
  }
}
Endpointhttps://YOUR-DOMAIN/mcp

Client configuration formats vary. Use this endpoint anywhere your agent accepts a remote Streamable HTTP MCP server.

01b Local

Or run it without the site.

The same read-only surface is available over stdio, reading the generated research artifacts straight from disk. It makes no network calls.

Local stdio server
npm run mcp:stdio

Point any stdio-capable MCP client at this command from the repository root. Useful in CI, offline, or when the Next.js server is not running.

02 Open formats

MCP is optional.

Agents that cannot connect to MCP can consume the same canonical records as concise Markdown or versioned JSON.

01

llms.txt

Compact research index for discovery and selective loading.

TXT ↗02

llms-full.txt

Complete curated export with evidence status, limitations, and sources.

TXT ↗03

Knowledge manifest

Schema version, freshness, record counts, and available interfaces.

JSON ↗04

Search API

Structured lexical retrieval across the normalized knowledge layer.

JSON ↗05

Canonical URLs

Every claim, project, pattern, and opportunity has a stable page matching its MCP resource URI.

XML ↗

03 Tools

Research before recommendation.

Tools return bounded structured results. Agents are instructed to retain claim status, separate hypotheses from observations, and propose validation before rollout.

  1. 01
    search_jev_knowledge

    Find claims, projects, patterns, opportunities, evidence, and synthesis with status preserved.

  2. 02
    get_jev_context_pack

    Retrieve a compact cited bundle for a goal without loading the entire corpus.

  3. 03
    assess_jev_fit

    Locate bounded semantic decisions, non-fits, deterministic boundaries, and validation steps.

  4. 04
    get_build_blueprint

    Open an authored opportunity with its architecture, MVP, experiment, and unknowns.

  5. 05
    trace_jev_claim

    Inspect evidence status, counterarguments, open questions, and original sources.

Trust contract

Context with provenance, not an oracle.

  • 01Official documentation remains authoritative for current API contracts.
  • 02Vendor claims stay labeled until independent evidence changes their status.
  • 03Authored opportunities are hypotheses to test, not guaranteed use cases.
  • 04Typed output constrains shape; it does not guarantee semantic correctness.
  • 05High-impact actions still require deterministic backstops and escalation.
  • 06A record’s website page, JSON export, and MCP resource resolve to one canonical address.
Read the complete human report →Last verified 18 Sept 2026