feat(aura-cli): aura runs registry — persist on sweep, list + rank (cycle D / iter 3)

Final iteration of the run-registry cycle (#33): the read surface that makes a
sweep family — and runs across separate invocations — comparable over time
(C18's "compare experiments over time", which has no home in git or Gitea).

- `aura sweep` now persists each point's RunReport to runs/runs.jsonl (append-
  only, one serde_json line per run) in addition to printing it.
- `aura runs list` prints every stored record, in store order.
- `aura runs rank <metric>` prints the stored runs best-first by the metric
  (total_pips desc; max_drawdown / exposure_sign_flips asc); an unknown metric
  is a usage error (exit 2), like every other aura arg error.

sweep_report splits into a production sweep_family() (the pure run) and a
#[cfg(test)] renderer kept for the in-bin goldens; the dispatch gains run_sweep
/ runs_list / runs_rank over default_registry() (runs/runs.jsonl under cwd).
Process goldens run the binary in a temp cwd so persistence never dirties the
repo; /runs/ is git-ignored as a backstop.

Verified end-to-end: two `aura sweep` invocations accumulate 8 records;
`aura runs list` shows all 8; `aura runs rank total_pips` orders them best-first;
`aura runs rank bogus` exits 2 with the known-metrics hint. cargo test
--workspace green (cli_run 11, registry 5, engine 93, …); clippy
--workspace --all-targets -D warnings clean; no stray runs/ in the work tree.

refs #33

crates/aura-cli/src/main.rs

@@ -11,8 +11,9 @@ mod render;
 use aura_core::{Firing, Scalar, ScalarKind, Timestamp};
 use aura_engine::{
     f64_field, summarize, sweep, BlueprintNode, Composite, Edge, FlatGraph, GridSpace, Harness,
-    OutField, ParamAlias, Role, RunManifest, RunReport, SourceSpec, Target,
+    OutField, ParamAlias, Role, RunManifest, RunReport, SourceSpec, SweepFamily, Target,
 };
+use aura_registry::{rank_by, Registry};
 use aura_std::{Ema, Exposure, Recorder, SimBroker, Sma, Sub};
 use std::sync::mpsc::{self, Receiver};
 
@@ -222,7 +223,7 @@ fn scalar_as_param_f64(s: &Scalar) -> f64 {
 /// the same report. Each point builds a fresh blueprint (fresh sink channels),
 /// bootstraps it under the point vector, runs it, and folds the drained sinks to
 /// metrics — the per-point closure the engine `sweep` drives disjointly.
-fn sweep_report() -> String {
+fn sweep_family() -> SweepFamily {
     let space = sample_blueprint_with_sinks().0.param_space();
     let grid = GridSpace::new(
         &space,
@@ -233,7 +234,7 @@ fn sweep_report() -> String {
         ],
     )
     .expect("the built-in grid matches the sample param-space");
-    let family = sweep(&grid, |point| {
+    sweep(&grid, |point| {
         let (bp, rx_eq, rx_ex) = sample_blueprint_with_sinks();
         let mut h = bp
             .bootstrap_with_params(point.to_vec())
@@ -261,15 +262,72 @@ fn sweep_report() -> String {
             },
             metrics: summarize(&equity, &exposure),
         }
-    });
+    })
+}
+
+/// Render a sweep family as one `RunReport` JSON line per point. Test helper:
+/// production (`run_sweep`) renders *and* persists per point.
+#[cfg(test)]
+fn sweep_report() -> String {
     let mut out = String::new();
-    for pt in &family.points {
+    for pt in &sweep_family().points {
         out.push_str(&pt.report.to_json());
         out.push('\n');
     }
     out
 }
 
+/// The default run registry: an append-only JSONL store under the current
+/// working directory. (A project-configured runs-dir via `Aura.toml` is a later
+/// refinement.)
+fn default_registry() -> Registry {
+    Registry::open("runs/runs.jsonl")
+}
+
+/// `aura sweep`: run the built-in sweep, persist each point's `RunReport` to the
+/// registry (the run record, queryable over time — C18), and print each as one
+/// JSON line.
+fn run_sweep() {
+    let reg = default_registry();
+    for pt in &sweep_family().points {
+        if let Err(e) = reg.append(&pt.report) {
+            eprintln!("aura: {e}");
+            std::process::exit(2);
+        }
+        println!("{}", pt.report.to_json());
+    }
+}
+
+/// `aura runs list`: print every stored run record, in store (over-time) order.
+fn runs_list() {
+    for report in &load_runs_or_exit() {
+        println!("{}", report.to_json());
+    }
+}
+
+/// `aura runs rank <metric>`: print the stored runs best-first by `metric`.
+fn runs_rank(metric: &str) {
+    match rank_by(load_runs_or_exit(), metric) {
+        Ok(ranked) => {
+            for report in &ranked {
+                println!("{}", report.to_json());
+            }
+        }
+        Err(e) => {
+            eprintln!("aura: {e}");
+            std::process::exit(2);
+        }
+    }
+}
+
+/// Load the registry or exit 2 with the error. A missing registry loads empty.
+fn load_runs_or_exit() -> Vec<RunReport> {
+    default_registry().load().unwrap_or_else(|e| {
+        eprintln!("aura: {e}");
+        std::process::exit(2);
+    })
+}
+
 // --- MACD proof-of-concept (a richer, nested indicator + strategy) -----------
 
 /// The MACD signal as a named composite: price → fast/slow `Ema` → the MACD line
@@ -426,7 +484,8 @@ fn run_macd() -> RunReport {
     }
 }
 
-const USAGE: &str = "usage: aura run [--macd] | aura graph | aura sweep";
+const USAGE: &str =
+    "usage: aura run [--macd] | aura graph | aura sweep | aura runs list | aura runs rank <metric>";
 
 fn main() {
     // Collect argv and match the whole vector: every accepted form is exhaustive,
@@ -437,7 +496,9 @@ fn main() {
         ["run"] => println!("{}", run_sample().to_json()),
         ["run", "--macd"] => println!("{}", run_macd().to_json()),
         ["graph"] => print!("{}", render::render_html(&sample_blueprint())),
-        ["sweep"] => print!("{}", sweep_report()),
+        ["sweep"] => run_sweep(),
+        ["runs", "list"] => runs_list(),
+        ["runs", "rank", metric] => runs_rank(metric),
         ["--help"] | ["-h"] => println!("{USAGE}"),
         _ => {
             eprintln!("aura: {USAGE}");