---
title: "V1 Get Field Orientation"
method: GET
path: "/v1/field-orientation"
tags: ["v1"]
---

# V1 Get Field Orientation

`GET /v1/field-orientation`

## Request

- Base URL: `https://api.scholarfeed.org` (the document declares no server; this is the origin it was published from)
- URL: `https://api.scholarfeed.org/v1/field-orientation`
- Auth: HTTP bearer

Cheap retrieval-only field orientation (TOOL-06, v3 get_field_orientation tool backend).

Embeds the topic via Gemini Flash (get_query_embedding), runs a cosine search
over the papers vector index to fetch candidate papers, then re-ranks those
candidates by a foundational-ness blend (0.6 * normalised citation_count +
0.4 * cosine_similarity). Returns the top `limit` papers as a thin list with
similarity score and citation count.

THIS IS NOT THE FIELD GUIDE AGENT. No DeepSeek synthesis, no quota, no Pro
requirement, no background jobs. The expensive LLM synthesis half stays in the
/field-guide skill (docs/architecture/tool-vs-skill.md §field_guide row) which
composes this retrieval endpoint + in-context reasoning client-side.

Rate limit: 20/minute per API key.

Response shape:
    topic: str
    papers: list[{arxiv_id, title, authors, published_date, citation_count,
                   rank_score, llm_summary, llm_novelty_score, similarity}]
    total: int
    note: str — reminder that this is retrieval-only, no LLM synthesis

## Query parameters

- `topic` string, required: Research area to orient on, e.g. 'diffusion models for protein structure'.
- `limit` integer: Number of foundational candidate papers to return (5–30).

## Response `200`

Successful Response

- object

## Other responses

- `422`: Validation Error

---

[API](https://skmtc.dev/scholarfeed/apis/scholar-feed-api.md) · [All operations](https://skmtc.dev/scholarfeed/apis/scholar-feed-api/llms.txt) · [OpenAPI document](https://skmtc.dev/scholarfeed/apis/scholar-feed-api/revisions/fc3200d27505?raw)