Search remote AI engineering jobs that use Claude Code, Cursor and other agentic coding tools.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Whole job descriptions for up to 25 postings named by id, in one call. Use this before judging anything — search_jobs summaries are too short to form an opinion on.
dedup_hash values from an earlier search.
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http agenticjobs https://api.agenticjobs.dev/mcp
Whole job descriptions for up to 25 postings named by id, in one call. Use this before judging anything — search_jobs summaries are too short to form an opinion on.
| Argument | Type | Description |
|---|---|---|
| ids* | array | dedup_hash values from an earlier search. |
{
"type": "object",
"required": [
"ids"
],
"properties": {
"ids": {
"type": "array",
"items": {
"type": "string"
},
"maxItems": 25,
"minItems": 1,
"description": "dedup_hash values from an earlier search."
}
}
}Single job with the full description and the LLM scoring reason.
| Argument | Type | Description |
|---|---|---|
| dedup_hash* | string |
{
"type": "object",
"required": [
"dedup_hash"
],
"properties": {
"dedup_hash": {
"type": "string"
}
}
}How to use this server: what the index holds, what the scores mean, what it will not do. Call this first, before any other tool.
| Argument | Type | Description |
|---|---|---|
| skill_version | integer | The version of the agenticjobs skill file you have saved, if you have one. |
{
"type": "object",
"properties": {
"skill_version": {
"type": "integer",
"description": "The version of the agenticjobs skill file you have saved, if you have one."
}
}
}Filter the agenticjobs.dev index by free-text and structured constraints. Returns up to `limit` rows sorted by vibe-coding fit, each description cut to 240 characters. The strongest single tool — call this first.
| Argument | Type | Description |
|---|---|---|
| q | string | Free-text — matches title, company, tags, AI-tool names. |
| ids | array | Narrow to these postings — pass ids from an earlier answer to re-check them under new conditions. |
| limit | integer | |
| tools | array | Restrict to jobs mentioning these AI-coding tools (e.g. Claude Code, Cursor, Codex). |
| source | array | |
| country | array | Full country names (Germany, UK, USA), not ISO codes. |
| min_score | integer | Minimum LLM rubric score: 70 is a good fit, 85 the curated top. |
| exclude_ids | array | Drop these postings — pass ids the person has already been shown. |
| remote_type | string |
Aggregate counts: total, good-matches, by source, plus when the index was last refreshed.
No arguments.
{
"type": "object",
"properties": {}
}Top N roles by vibe-coding fit — use when the user just wants the best matches without supplying filters.
| Argument | Type | Description |
|---|---|---|
| limit | integer |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 20,
"maximum": 50,
"minimum": 1
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 11, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.
{
"type": "object",
"properties": {
"q": {
"type": "string",
"description": "Free-text — matches title, company, tags, AI-tool names."
},
"ids": {
"type": "array",
"items": {
"type": "string"
},
"description": "Narrow to these postings — pass ids from an earlier answer to re-check them under new conditions."
},
"limit": {
"type": "integer",
"default": 50,
"maximum": 200,
"minimum": 1
},
"tools": {
"type": "array",
"items": {
"type": "string"
},
"description": "Restrict to jobs mentioning these AI-coding tools (e.g. Claude Code, Cursor, Codex)."
},
"source": {
"type": "array",
"items": {
"type": "string"
}
},
"country": {
"type": "array",
"items": {
"type": "string"
},
"description": "Full country names (Germany, UK, USA), not ISO codes."
},
"min_score": {
"type": "integer",
"maximum": 100,
"minimum": 0,
"description": "Minimum LLM rubric score: 70 is a good fit, 85 the curated top."
},
"exclude_ids": {
"type": "array",
"items": {
"type": "string"
},
"description": "Drop these postings — pass ids the person has already been shown."
},
"remote_type": {
"enum": [
"fully_remote",
"hybrid",
"onsite"
],
"type": "string"
}
}
}