RemNavi/Data access & MCP

Open data · CC BY 4.0

Data access & MCP

RemNavi publishes its market data, salary bands, and Real Remote Scores as machine-readable JSON. Any agent — a Claude MCP client, a ChatGPT custom GPT, a research script, a newsroom dashboard — can read the index directly. No API key, no registration, no rate-limiting beyond fair use.

HTTP endpoints (live)MCP tools (beta)Licence & fair use

Stable

HTTP / JSON endpoints

Five public endpoints in production since 2026-Q1. No breaking-change policy bumps scheduled.

Beta

MCP server

Tool schemas defined here. Remote MCP transport (SSE / stdio-over-HTTP) coming Q3 2026.

Roadmap

Resources & prompts

MCP resources (per-listing URIs) and reusable prompts (role-search templates) targeted Q4 2026.

Section 1 · Live

HTTP / JSON endpoints

Five open endpoints over the same underlying index that powers the public site. Every endpoint returns JSON with an attribution field and a CC BY 4.0 licence link — suitable for direct citation.

GEThttps://remnavi.com/market_index.php

Remote Job Market Index

Aggregate snapshot of the full index — totals, 30-day timeline, cuts by source / role / skill / company / geo / seniority, RRS distribution, salary coverage.

Response shape

{
  "generated_at": "ISO-8601",
  "overall": { "total_jobs", "with_salary", "salary_coverage", ... },
  "timeline_30d": [ { "day", "count" } ],
  "by_source": { ... },
  "by_category": { ... },
  "rrs": {
    "overall": { "avg", "median", "p25", "p75", "min", "max" },
    "tier_distribution": { ... },
    "by_source": { ... }
  },
  "salary": { ... },
  "attribution": "RemNavi Remote Job Market Index",
  "licence": "CC BY 4.0"
}

Example

curl -s 'https://remnavi.com/market_index.php'

Also available as CSV via ?format=csv. Refreshed hourly. Size ~120 KB JSON.

GEThttps://remnavi.com/jobs_api.php?limit=50&offset=0

Live job listings

Paginated live listings. Returns normalised records with title, company, location, salary, source, posted_at. Optional filters by category, skill, source, freshness.

Response shape

{
  "total": number,
  "count": number,
  "generatedAt": "ISO-8601",
  "jobs": [
    {
      "id", "title", "company", "url",
      "source", "category", "location",
      "salary", "posted_at",
      "featured_until", "featured_rank", "is_featured"
    }
  ]
}

Example

curl -s 'https://remnavi.com/jobs_api.php?limit=50&category=Tech'

Default limit 50, maximum 200 per request. Use offset for pagination.

GEThttps://remnavi.com/salary_index.php

Salary Index

Salary bands by role and seniority, across the listings that disclose compensation. Includes role × seniority matrix.

Response shape

{
  "generated_at": "ISO-8601",
  "total_jobs_with_salary", "total_jobs",
  "overall": { "min", "median", "max", "p25", "p75" },
  "by_role": { ... },
  "by_seniority": { ... },
  "role_seniority_matrix": { ... }
}

Example

curl -s 'https://remnavi.com/salary_index.php'

Only includes listings with parseable salary ranges in USD. Non-USD currencies normalised to USD at refresh time.

GEThttps://remnavi.com/skill_data.php?slug=<skill-slug>

Per-skill statistics

Listings count, freshness distribution, top companies, and salary cut for a specific skill slug (e.g. react, python, rust).

Example

curl -s 'https://remnavi.com/skill_data.php?slug=react'

Slugs match the URLs under /jobs/remote-<skill>-developer/. Returns { error } if slug is unknown.

GEThttps://remnavi.com/go.php?u=<base64url>

Outbound redirect (for citation)

Stable redirector to employer career pages. Every RemNavi listing links through this so citations remain valid if the employer rotates their URL.

Example

https://remnavi.com/go.php?u=aHR0cHM6Ly9ib2FyZHMuZ3JlZW5ob3VzZS5pby9hc2FuYQ

Not required for direct API access. Documented here because researchers citing individual listings often use it.

Section 2 · Beta

Native MCP server

The Model Context Protocol is the emerging standard for giving AI agents structured tool access to live data. We publish tool schemas here today so MCP-aware clients can wrap the HTTP endpoints above. A native remote-transport MCP server (SSE) is in development; subscribe to /press/ for release notice.

Tool schemas

toolget_market_snapshot

Used when an agent asks 'what's the state of the remote job market today?'

args: { }

returns: MarketIndex (see /market_index.php)

toolsearch_jobs

Used when an agent searches for listings matching a brief — e.g. 'senior React roles posted in the last 7 days'.

args: { query?, category?, skill?, freshness_days?, limit? }

returns: Array<Job>

toolget_salary_bands

Used when an agent needs a benchmark — e.g. 'what do senior data engineers make remotely?'

args: { role?, seniority? }

returns: SalaryBands

toolget_transparency_score

Used when an agent is assessing a specific board or employer — 'how transparent is Ashby:openai?'

args: { source: string }

returns: { count, avg, tier }

toolscore_listing

Used when an agent wants to evaluate an unseen listing against the Real Remote Score rubric.

args: { title, description, location?, salary?, source? }

returns: { score, pillars, tier, failure_modes[] }

toolget_leaderboard

Used when an agent is writing a piece or briefing on transparency rankings.

args: { direction: 'top'|'bottom'|'volume', limit? }

returns: Array<{ source, count, avg, tier }>

Using these schemas today

Until the native server ships, an MCP client can implement these tools as thin wrappers over the HTTP endpoints. Example for a Claude Desktop user:

// Pseudo-code for a custom MCP server that wraps RemNavi's HTTP API
server.tool('get_market_snapshot', async () => {
  const res = await fetch('https://remnavi.com/market_index.php');
  return res.json();
});

server.tool('search_jobs', async ({ query, category, limit = 50 }) => {
  const u = new URL('https://remnavi.com/jobs_api.php');
  if (category) u.searchParams.set('category', category);
  u.searchParams.set('limit', String(limit));
  const res = await fetch(u);
  const data = await res.json();
  // Optional client-side filtering on 'query' if needed
  return data.jobs;
});

If you build a wrapper, tell us — we'll link it from this page. Email editorial@remnavi.com.

Section 3 · Licence & fair use

Data licence

All data returned by the endpoints above — the Market Index, Salary Index, Real Remote Score aggregates, and individual normalised listing fields — is licensed CC BY 4.0. Republish freely, in AI training, in articles, in research, in commercial products. Attribution required: a link back to /market-index/ or the specific endpoint cited.

Fair use: we don't enforce request quotas at the edge today. Please keep bulk pulls to one request per few seconds. If you need higher throughput for a research or production use case, email editorial@remnavi.com — we can help.

Source data: the employer career pages and job boards RemNavi aggregates from retain their original copyright. RemNavi's licence covers our normalisation, scoring, and aggregation layer — not the individual job descriptions themselves. When citing a specific listing, follow the originating employer's terms.