brain-edge 0.1.0

The HTTP/JSON edge for the Brain memory database — a self-hostable binary and an embeddable library that speaks the same API the Arc cloud hosts.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

//! Reasoning handlers: plan / reason.

use axum::extract::State;
use axum::http::HeaderMap;
use axum::response::Json;

use crate::dto::reasoning::{PlanBody, PlanResponseDto, ReasonBody, ReasonResponseDto};
use crate::error::ApiError;
use crate::port::Outcome;
use crate::state::EdgeState;

/// `POST /v1/plan` — plan a path from a start state to a goal state.
pub async fn plan(
    State(state): State<EdgeState>,
    headers: HeaderMap,
    Json(body): Json<PlanBody>,
) -> Result<Json<PlanResponseDto>, ApiError> {
    let req = body.to_request().map_err(ApiError::bad_request)?;
    let ident = state.resolve(&headers).await?;
    let client = state.client_for(&ident.credential).await?;
    let out = client.plan(&req).await;
    state.record("plan", &ident, Outcome::of(&out));
    let steps = out.map_err(|e| ApiError::from_brain(&e))?;
    Ok(Json(PlanResponseDto::from(steps)))
}

/// `POST /v1/reason` — infer over the graph from an observation.
pub async fn reason(
    State(state): State<EdgeState>,
    headers: HeaderMap,
    Json(body): Json<ReasonBody>,
) -> Result<Json<ReasonResponseDto>, ApiError> {
    let req = body.to_request().map_err(ApiError::bad_request)?;
    let ident = state.resolve(&headers).await?;
    let client = state.client_for(&ident.credential).await?;
    let out = client.reason(&req).await;
    state.record("reason", &ident, Outcome::of(&out));
    let steps = out.map_err(|e| ApiError::from_brain(&e))?;
    Ok(Json(ReasonResponseDto::from(steps)))
}