> ## Documentation Index
> Fetch the complete documentation index at: https://retrievalcontextprotocol.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Feedback

> The one client→server channel — relevance, reward, and integrity signals that RL-trained and corrective retrievers consume.

`feedback` is gated by **`feedback`**. Every other method flows a request from
client to server and an answer back; `feedback` is the reverse — it tells the
server how earlier hits *fared downstream* so an RL-trained retriever
(Search-R1, DeepRetrieval) can learn, a corrective loop can mark a hit
used/unhelpful, and security tooling can flag suspected poisoning.

It is **side-effect-only**: a `feedback` call **MUST NOT** change the result of
any concurrent `retrieve`. What the server does with the signal — log, online
update, quarantine — is its choice.

<CodeGroup>
  ```json Request theme={null}
  {
    "jsonrpc": "2.0", "id": 1, "method": "feedback",
    "params": {
      "sessionId": "traj-42",
      "query": "who invented the transformer?",
      "signals": [
        { "hitId": "arxiv:1706.03762#3", "used": true, "cited": true, "reward": 0.8 },
        { "hitId": "blog:spam#1", "poisonSuspected": true }
      ]
    }
  }
  ```

  ```json Response theme={null}
  { "jsonrpc": "2.0", "id": 1, "result": { "accepted": 2 } }
  ```
</CodeGroup>

## Params

<ParamField path="signals" type="array" required>
  One or more per-hit signal objects.
</ParamField>

<ParamField path="sessionId" type="string">
  Ties the feedback to an [agentic trajectory](/concepts/agentic-rag).
</ParamField>

<ParamField path="query" type="string">
  The query the hits answered (helps the server attribute the signal).
</ParamField>

### Signal object

<ParamField path="hitId" type="string | number" required>
  The `id` of the hit the signal is about.
</ParamField>

<ParamField path="used" type="boolean">
  The hit was placed in the final context.
</ParamField>

<ParamField path="cited" type="boolean">
  The hit was cited in the generated answer.
</ParamField>

<ParamField path="helpful" type="boolean">
  A human/judge relevance label.
</ParamField>

<ParamField path="reward" type="number">
  A scalar reward in `[-1, 1]` for RL-trained retrievers.
</ParamField>

<ParamField path="poisonSuspected" type="boolean">
  Integrity signal — the hit is suspected corpus poisoning.
</ParamField>

<ParamField path="injectionSuspected" type="boolean">
  Safety signal — the hit is suspected to carry an indirect prompt injection.
</ParamField>

## Result

`{ "accepted": <int> }` — how many signals the server recorded. A server **MAY**
ignore any field it does not act on (still counting it) and **MUST** treat every
signal as **untrusted, advisory input** — never as a control instruction, and
never as authorisation to act against other tenants.

<Info>
  Feedback closes the loop for reinforcement-learning retrieval (reward → policy
  update) and corrective RAG (used/helpful → re-rank or re-index), and gives
  security tooling a path to report poisoning so the server can quarantine a
  source. It is entirely optional: a client that never sends it and a server that
  only logs it are both fully conforming.
</Info>