Aura/docs/plans/0029-run-registry-cli.md

# Run Registry — Iteration 3 (CLI persistence + read surface) — Implementation Plan

> **Parent spec:** `docs/specs/0029-run-registry-sweep-family.md`
>
> **For agentic workers:** REQUIRED SUB-SKILL: use the `implement` skill to run
> this plan. Steps use `- [ ]` checkboxes for tracking.

**Goal:** Wire `aura-registry` into the CLI: `aura sweep` persists each point's
`RunReport` to `runs/runs.jsonl`; `aura runs list` prints every stored record;
`aura runs rank <metric>` prints them best-first; with process-level goldens.

**Architecture:** `sweep_report` is split — a production `sweep_family() ->
SweepFamily` (the pure run) and a `#[cfg(test)] sweep_report() -> String`
renderer kept for the in-bin tests. The `["sweep"]` arm becomes `run_sweep`,
which persists each point via `default_registry()` and prints it. Two new arms,
`["runs","list"]` and `["runs","rank", metric]`, read the registry back (typed,
via serde) and print via `RunReport::to_json`. Process tests run the binary in a
temp cwd so persistence never dirties the repo; `/runs/` is git-ignored as a
backstop.

**Tech Stack:** `aura-cli`, `aura-registry` (Registry, rank_by).

---

**Files this plan creates or modifies:**

- Modify: `.gitignore` — ignore `/runs/`
- Modify: `crates/aura-cli/Cargo.toml:12-15` — add the `aura-registry` dependency
- Modify: `crates/aura-cli/src/main.rs:12-15` — import `SweepFamily` + `aura_registry::{rank_by, Registry}`
- Modify: `crates/aura-cli/src/main.rs:225-271` — split `sweep_report` into `sweep_family` + `#[cfg(test)] sweep_report`; add `default_registry`/`run_sweep`/`runs_list`/`runs_rank`/`load_runs_or_exit`
- Modify: `crates/aura-cli/src/main.rs:429` — extend `USAGE`
- Modify: `crates/aura-cli/src/main.rs:440` — replace the `["sweep"]` arm + add `runs` arms
- Test: `crates/aura-cli/tests/cli_run.rs` — temp-cwd helper, update the sweep golden, add the runs-flow / empty / strictness tests

---

### Task 1: Prep — git-ignore `/runs/` and add the registry dependency

**Files:**
- Modify: `.gitignore`
- Modify: `crates/aura-cli/Cargo.toml:12-15`

- [ ] **Step 1: Ignore the default registry path**

In `.gitignore`, append:

```
# `aura sweep` persists run records to ./runs/runs.jsonl by default; that store
# is local run telemetry, not a repo artifact.
/runs/
```

- [ ] **Step 2: Add the aura-registry dependency**

In `crates/aura-cli/Cargo.toml`, replace the `[dependencies]` block (lines
12-15):

```toml
[dependencies]
aura-core = { path = "../aura-core" }
aura-engine = { path = "../aura-engine" }
aura-std = { path = "../aura-std" }
```

with:

```toml
[dependencies]
aura-core = { path = "../aura-core" }
aura-engine = { path = "../aura-engine" }
aura-registry = { path = "../aura-registry" }
aura-std = { path = "../aura-std" }
```

- [ ] **Step 3: Verify the workspace still builds**

Run: `cargo build --workspace`
Expected: builds green, exit 0 (the new dependency is not yet used — that is
fine; `unused_crate_dependencies` is allow-by-default).

---

### Task 2: CLI wiring — persist on sweep, list + rank subcommands (RED-first)

**Files:**
- Modify: `crates/aura-cli/tests/cli_run.rs`
- Modify: `crates/aura-cli/src/main.rs`

- [ ] **Step 1: Write the failing process tests**

In `crates/aura-cli/tests/cli_run.rs`, add the temp-cwd helper just after the
`BIN` constant (line 8):

```rust
/// A fresh, unique working directory for a process test that persists run
/// records (so `aura sweep`'s `./runs/runs.jsonl` never dirties the repo).
/// Unique per test + per process; no external tempfile dependency.
fn temp_cwd(name: &str) -> std::path::PathBuf {
    let dir = std::env::temp_dir().join(format!("aura-cli-{}-{}", std::process::id(), name));
    let _ = std::fs::remove_dir_all(&dir);
    std::fs::create_dir_all(&dir).expect("create temp cwd");
    dir
}
```

Then replace `sweep_prints_four_json_lines_and_exits_zero` (the body up to and
including its closing `}`) — it must now run in a temp cwd, since `aura sweep`
persists:

```rust
#[test]
fn sweep_prints_four_json_lines_and_exits_zero() {
    let cwd = temp_cwd("sweep4");
    let out = Command::new(BIN).arg("sweep").current_dir(&cwd).output().expect("spawn aura sweep");
    assert!(out.status.success(), "exit status: {:?}", out.status);

    let stdout = String::from_utf8(out.stdout).expect("utf-8 stdout");
    assert_eq!(stdout.lines().count(), 4, "stdout was: {stdout:?}");
    let first = stdout.lines().next().unwrap();
    // each line is a full RunReport; commit is the real git HEAD (volatile), so
    // pin the first odometer point's manifest params, not the commit value.
    assert!(first.starts_with("{\"manifest\":{\"commit\":\""), "got: {stdout}");
    assert!(
        first.contains("\"params\":{\"sma_cross.fast\":2,\"sma_cross.slow\":4,\"scale\":0.5}"),
        "got: {stdout}"
    );
    let _ = std::fs::remove_dir_all(&cwd);
}
```

Then add the new tests at the end of the file:

```rust
#[test]
fn runs_list_and_rank_across_invocations() {
    let cwd = temp_cwd("runs-flow");
    // two sweeps accumulate 8 records over time
    for _ in 0..2 {
        let out = Command::new(BIN).arg("sweep").current_dir(&cwd).output().expect("spawn sweep");
        assert!(out.status.success(), "sweep exit: {:?}", out.status);
    }

    // `runs list` shows all 8 stored records
    let list = Command::new(BIN).args(["runs", "list"]).current_dir(&cwd).output().expect("spawn list");
    assert!(list.status.success(), "list exit: {:?}", list.status);
    let list_out = String::from_utf8(list.stdout).expect("utf-8");
    assert_eq!(list_out.lines().count(), 8, "list: {list_out:?}");
    assert!(list_out.lines().all(|l| l.starts_with("{\"manifest\":{")), "list shape: {list_out:?}");

    // `runs rank total_pips`: 8 lines, highest total_pips first
    let rank = Command::new(BIN).args(["runs", "rank", "total_pips"]).current_dir(&cwd).output().expect("spawn rank");
    assert!(rank.status.success(), "rank exit: {:?}", rank.status);
    let rank_out = String::from_utf8(rank.stdout).expect("utf-8");
    assert_eq!(rank_out.lines().count(), 8, "rank: {rank_out:?}");
    let pips: Vec<f64> = rank_out
        .lines()
        .map(|l| {
            let key = "\"total_pips\":";
            let start = l.find(key).expect("line has total_pips") + key.len();
            let tail = &l[start..];
            let end = tail.find([',', '}']).expect("token end");
            tail[..end].parse().expect("total_pips is an f64")
        })
        .collect();
    assert!(pips.windows(2).all(|w| w[0] >= w[1]), "rank not descending: {pips:?}");

    // unknown metric is a usage error
    let bogus = Command::new(BIN).args(["runs", "rank", "bogus"]).current_dir(&cwd).output().expect("spawn bogus");
    assert_eq!(bogus.status.code(), Some(2), "bogus exit: {:?}", bogus.status);

    let _ = std::fs::remove_dir_all(&cwd);
}

#[test]
fn runs_list_on_empty_registry_is_zero_lines_exit_zero() {
    let cwd = temp_cwd("runs-empty");
    let out = Command::new(BIN).args(["runs", "list"]).current_dir(&cwd).output().expect("spawn list");
    assert!(out.status.success(), "list exit: {:?}", out.status);
    assert_eq!(out.stdout.len(), 0, "expected no output, got: {:?}", out.stdout);
    let _ = std::fs::remove_dir_all(&cwd);
}

#[test]
fn runs_bare_subcommand_is_strict_and_exits_two() {
    // `aura runs` with no/unknown sub-subcommand is a usage error (#16 strict).
    let out = Command::new(BIN).arg("runs").output().expect("spawn aura runs");
    assert_eq!(out.status.code(), Some(2), "exit status: {:?}", out.status);
}
```

- [ ] **Step 2: Run the new tests to verify they fail**

Run: `cargo test -p aura-cli --test cli_run runs_`
Expected: FAIL — `runs_list_and_rank_across_invocations` and
`runs_list_on_empty_registry_is_zero_lines_exit_zero` fail because the current
binary has no `runs` arm, so `aura runs list` falls through to the usage-error
path and exits 2 (the success/line-count assertions fail).
(`runs_bare_subcommand_is_strict_and_exits_two` already passes — `aura runs` is
already a usage error today.)

- [ ] **Step 3: Add the imports**

In `crates/aura-cli/src/main.rs`, replace the import block (lines 12-15):

```rust
use aura_engine::{
    f64_field, summarize, sweep, BlueprintNode, Composite, Edge, FlatGraph, GridSpace, Harness,
    OutField, ParamAlias, Role, RunManifest, RunReport, SourceSpec, Target,
};
use aura_std::{Ema, Exposure, Recorder, SimBroker, Sma, Sub};
```

with (add `SweepFamily` to the engine list; add the registry import):

```rust
use aura_engine::{
    f64_field, summarize, sweep, BlueprintNode, Composite, Edge, FlatGraph, GridSpace, Harness,
    OutField, ParamAlias, Role, RunManifest, RunReport, SourceSpec, SweepFamily, Target,
};
use aura_registry::{rank_by, Registry};
use aura_std::{Ema, Exposure, Recorder, SimBroker, Sma, Sub};
```

- [ ] **Step 4: Split `sweep_report` into `sweep_family` + a test renderer**

In `crates/aura-cli/src/main.rs`, replace the whole `sweep_report` function
(from `fn sweep_report() -> String {` at line 225 through its closing `}` at
line 271 — keep the doc comment at line 224 above it):

```rust
fn sweep_report() -> String {
    let space = sample_blueprint_with_sinks().0.param_space();
    let grid = GridSpace::new(
        &space,
        vec![
            vec![Scalar::I64(2), Scalar::I64(3)], // fast  ∈ {2, 3}
            vec![Scalar::I64(4), Scalar::I64(5)], // slow  ∈ {4, 5}
            vec![Scalar::F64(0.5)],               // scale ∈ {0.5}
        ],
    )
    .expect("the built-in grid matches the sample param-space");
    let family = sweep(&grid, |point| {
        let (bp, rx_eq, rx_ex) = sample_blueprint_with_sinks();
        let mut h = bp
            .bootstrap_with_params(point.to_vec())
            .expect("grid points are kind-checked against param_space");
        let prices = synthetic_prices();
        let window = (
            prices.first().expect("non-empty stream").0,
            prices.last().expect("non-empty stream").0,
        );
        h.run(vec![prices]);
        let equity = f64_field(&rx_eq.try_iter().collect::<Vec<_>>(), 0);
        let exposure = f64_field(&rx_ex.try_iter().collect::<Vec<_>>(), 0);
        let params = space
            .iter()
            .zip(point)
            .map(|(ps, v)| (ps.name.clone(), scalar_as_param_f64(v)))
            .collect();
        RunReport {
            manifest: RunManifest {
                commit: option_env!("AURA_COMMIT").unwrap_or("unknown").to_string(),
                params,
                window,
                seed: 0,
                broker: "sim-optimal(pip_size=0.0001)".to_string(),
            },
            metrics: summarize(&equity, &exposure),
        }
    });
    let mut out = String::new();
    for pt in &family.points {
        out.push_str(&pt.report.to_json());
        out.push('\n');
    }
    out
}
```

with (the run becomes the production `sweep_family`; the renderer stays as a
`#[cfg(test)]` helper so it is not dead code in a normal build):

```rust
fn sweep_family() -> SweepFamily {
    let space = sample_blueprint_with_sinks().0.param_space();
    let grid = GridSpace::new(
        &space,
        vec![
            vec![Scalar::I64(2), Scalar::I64(3)], // fast  ∈ {2, 3}
            vec![Scalar::I64(4), Scalar::I64(5)], // slow  ∈ {4, 5}
            vec![Scalar::F64(0.5)],               // scale ∈ {0.5}
        ],
    )
    .expect("the built-in grid matches the sample param-space");
    sweep(&grid, |point| {
        let (bp, rx_eq, rx_ex) = sample_blueprint_with_sinks();
        let mut h = bp
            .bootstrap_with_params(point.to_vec())
            .expect("grid points are kind-checked against param_space");
        let prices = synthetic_prices();
        let window = (
            prices.first().expect("non-empty stream").0,
            prices.last().expect("non-empty stream").0,
        );
        h.run(vec![prices]);
        let equity = f64_field(&rx_eq.try_iter().collect::<Vec<_>>(), 0);
        let exposure = f64_field(&rx_ex.try_iter().collect::<Vec<_>>(), 0);
        let params = space
            .iter()
            .zip(point)
            .map(|(ps, v)| (ps.name.clone(), scalar_as_param_f64(v)))
            .collect();
        RunReport {
            manifest: RunManifest {
                commit: option_env!("AURA_COMMIT").unwrap_or("unknown").to_string(),
                params,
                window,
                seed: 0,
                broker: "sim-optimal(pip_size=0.0001)".to_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 &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);
    })
}
```

- [ ] **Step 5: Extend USAGE and the dispatch arms**

In `crates/aura-cli/src/main.rs`, replace the `USAGE` constant (line 429):

```rust
const USAGE: &str = "usage: aura run [--macd] | aura graph | aura sweep";
```

with:

```rust
const USAGE: &str =
    "usage: aura run [--macd] | aura graph | aura sweep | aura runs list | aura runs rank <metric>";
```

Then replace the `["sweep"]` dispatch arm (line 440):

```rust
        ["sweep"] => print!("{}", sweep_report()),
```

with:

```rust
        ["sweep"] => run_sweep(),
        ["runs", "list"] => runs_list(),
        ["runs", "rank", metric] => runs_rank(metric),
```

- [ ] **Step 6: Run the new tests to verify they pass**

Run: `cargo test -p aura-cli --test cli_run runs_`
Expected: PASS — `runs_list_and_rank_across_invocations`,
`runs_list_on_empty_registry_is_zero_lines_exit_zero`, and
`runs_bare_subcommand_is_strict_and_exits_two` all green.

- [ ] **Step 7: Full workspace gate**

Run: `cargo test --workspace`
Expected: PASS — all crates green, including the updated `aura sweep` golden and
the in-bin `sweep_report_*` tests (still calling the `#[cfg(test)] sweep_report`).

Run: `cargo clippy --workspace --all-targets -- -D warnings`
Expected: no warnings (no dead `sweep_report`; no unused imports).