agent-skills

# JobYap agent tools

Connect any AI agent to [JobYap](https://jobyap.com) — live job postings aggregated from
company career sites, each with salaries, full descriptions and a community discussion
thread. Everything is read-only and needs no account or API key.

Two ways in, use either or both:

- **MCP server** (remote, streamable HTTP): `https://mcp.jobyap.com/mcp`
- **Skill**: [`skills/jobyap`](skills/jobyap/SKILL.md) teaches agents the JSON API
  (`https://jobyap.com/api/v1`) with curl — works even without MCP support.

## Install the MCP server

**Claude Code**

```bash
claude mcp add --transport http jobyap https://mcp.jobyap.com/mcp
```

**Claude.ai / Claude Desktop** — Settings → Connectors → Add custom connector →
`https://mcp.jobyap.com/mcp` (no authentication).

**Cursor** — click the deeplink, or add the config manually:

[Add to Cursor](cursor://anysphere.cursor-deeplink/mcp/install?name=jobyap&config=eyJ1cmwiOiJodHRwczovL21jcC5qb2J5YXAuY29tL21jcCJ9)

```json
{ "mcpServers": { "jobyap": { "url": "https://mcp.jobyap.com/mcp" } } }
```

**VS Code / GitHub Copilot**

[Install in VS Code](vscode:mcp/install?%7B%22name%22%3A%22jobyap%22%2C%22type%22%3A%22http%22%2C%22url%22%3A%22https%3A%2F%2Fmcp.jobyap.com%2Fmcp%22%7D)

```json
{ "servers": { "jobyap": { "type": "http", "url": "https://mcp.jobyap.com/mcp" } } }
```

**ChatGPT** — Settings → Connectors (developer mode) → Add →
`https://mcp.jobyap.com/mcp`, No Authentication. The server implements the deep-research
`search`/`fetch` contract.

**Goose**

[Add to Goose](goose://extension?url=https%3A%2F%2Fmcp.jobyap.com%2Fmcp&type=streamable_http&id=jobyap&name=JobYap)

**Windsurf / Zed / JetBrains / LM Studio / others** — any streamable-HTTP MCP client
works; point it at `https://mcp.jobyap.com/mcp`.

**stdio-only clients** — use the [`jobyap-mcp`](https://www.npmjs.com/package/jobyap-mcp)
bridge (wraps [mcp-remote](https://www.npmjs.com/package/mcp-remote)):

```json
{ "mcpServers": { "jobyap": { "command": "npx", "args": ["-y", "jobyap-mcp"] } } }
```

## Install the skill

**Cross-agent** (Claude Code, Cursor, Codex, Gemini-compatible CLIs, opencode and ~70
others) via the [skills CLI](https://skills.sh):

```bash
npx skills add jobyap/agent-skills
```

**Claude Code plugin** — installs the skill *and* registers the MCP server in one go.
Two steps: add the marketplace, then install the plugin:

```
/plugin marketplace add jobyap/agent-skills
/plugin install jobyap@jobyap
```

**Manual** — copy `skills/jobyap/` into your agent's skills directory
(`.claude/skills/`, `.cursor/skills/`, `.agents/skills/`, …).

## Google Antigravity

The [`antigravity/jobyap`](antigravity/jobyap) plugin bundles the MCP server config
(`serverUrl`) and the skill:

```bash
git clone https://github.com/jobyap/agent-skills && agy plugin install ./agent-skills/antigravity/jobyap
```

Or add the MCP server manually with `serverUrl: https://mcp.jobyap.com/mcp`.

## What the tools do

| Tool | Purpose |
| --- | --- |
| `search_jobs` | Filtered search: title keywords, company, location identifiers, remote/hybrid, freshness, popularity; cursor pagination. |
| `get_job` | One posting in full: markdown description, salaries, locations, employment types, apply URL. Accepts ids or jobyap.com URLs. |
| `get_job_comments` | The community discussion thread (threaded, chronological, like counts). |
| `search_locations` | Free text → location filter identifiers. |
| `list_companies` | Tracked companies with active-job counts. |
| `get_job_stats` | Companies tracked, active jobs, newest posting date. |
| `search` / `fetch` | OpenAI deep-research connector contract over the same data. |

Full API reference: [skills/jobyap/references/api.md](skills/jobyap/references/api.md) ·
Docs: [jobyap.com/agents](https://jobyap.com/agents) · [llms.txt](https://jobyap.com/llms.txt)