29bf6015585d8f6c5c06dfea1e2536bbfb1decaa
97 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Brummel
|
fc9cf23b87 |
feat(cli,docs): ship the ganged open examples + document the gang construct (#61 tasks 7-8)
The two open examples now expose their author-intended single knobs — the closed builders always bound these pairs to one value by hand; the open forms falsely offered them as independent axes: - r_breakout_open: channel_hi.length + channel_lo.length -> channel_length (the Donchian channel is structurally ONE parameter) - r_meanrev_open: mean_window.length + var_window.length -> window; the band factor stays an independent axis Regenerated via the emitters (never hand-edited); the closed examples are byte-unchanged. The carve builders gang in the open branch only. Test migration: the two axis-namespace pins, five --real e2e invocations (single gang axis, the 10,20 diagonal for the campaign pair — the mismatched 20,40/10,20 grid was exactly the configuration space the gang retires), and the param_stability row counts (4 -> 3). The r-sma walkforward golden anchor is untouched (it never swept a ganged pair). Docs: authoring-guide gains the seventh op + the third param state (open/bound/ganged) + the gang wrap note; README op list + Axis concept; glossary gang entry; ledger C24 records the gang verb and the pre-ship Tier-2 dormancy (no format-version bump while no out-of-repo reader exists). Verified: full workspace suite 1104/0 (--real e2e included, local data present), clippy -D warnings clean, cargo doc clean; live acceptance: introspect --params prints channel_length:I64 alone / window:I64 + band.factor:F64. closes #61 |
||
|
Brummel
|
692760d3b0 |
test(cli): RED — aura graph with a bad blueprint arg must be a usage fault (#28)
The #28 cycle-close audit found the just-shipped `aura graph <file>` silently renders the embedded sample at exit 0 when the positional names an unreadable blueprint (typo, nonexistent path, wrong extension): is_blueprint_file returns None both for "no positional" (the legit sample default) and for a bad arg, and dispatch_graph's None arm conflates them — re-introducing the exact #28 "I asked for my graph, got the sample" surprise in the error path. This pins the fix: a named-but-unreadable arg exits 2 like its siblings, while bare `aura graph` still renders the sample. refs #28 |
||
|
Brummel
|
7cc986486f |
test(cli): RED — aura graph must render a consumer blueprint, not the sample (#28)
aura graph is the last verb still sample-bound: its no-subcommand arm always renders the embedded r_sma_open.json, ignoring any blueprint path, while run/sweep/mc/walkforward/generalize consume a consumer blueprint file via is_blueprint_file. This pins the target: aura graph <blueprint.json> renders that file's own graph (its inlined model reflects the file's nodes, e.g. r-breakout's channel_hi), not the embedded sample. refs #28 |
||
|
Brummel
|
94cdbf90cc |
feat(campaign,cli): permit --select plateau in the walk_forward stage (#215)
The campaign walk_forward stage picked its per-window in-sample winner by optimize_deflated (argmax) unconditionally, dropping the stage's `select` field, and both the preflight and the `walkforward` verb's `--real` campaign sub-branch refused plateau:* with a forward-pointer here (the #210 decision-4 deferral; the synthetic non-`--real` sub-branch has accepted plateau since #173). Honour plateau selection when no gate precedes the wf stage. - Preflight: the plateau-in-wf refusal is narrowed to gate-preceded stages. With no preceding gate the survivors are the full sweep grid, so the parameter lattice (grid.axis_lens) is exact and optimize_plateau is sound; a gate filters survivors below the grid, breaking the lattice, so plateau after any gate stays refused. The refusal is static — document validity must be data-independent (a gate's filtering depends on the data) — and keeps the existing "a gated survivor subset has no parameter lattice" message, which now describes exactly the only case it fires. - Executor: run_walk_forward_stage threads the per-cell grid.axis_lens and picks each window's in-sample winner via select_sweep_winner — the same dispatch the sweep stage uses. (Argmax, deflate=true) is byte-identical to the old optimize_deflated call, so the argmax grade and content-id pins are unchanged; the plateau arms read axis_lens and ignore deflate. - CLI: the walkforward verb threads --select through translate_walkforward / run_walkforward_sugar (a walkforward-local param, not the shared four-verb SugarInvocation), mapping the parsed Selection to the research SelectRule via select_rule_of; the refusal and its select_is_plateau helper are deleted. Scope is walkforward-only — the mc verb's wf-stage select is untouched. The plateau:worst grade anchor (stitched_total_pips = -9683776.67) is distinct from the argmax anchor (-10398606.67), proving the plateau path genuinely changes winner selection rather than falling back to argmax. A gate-free plateau process validates clean while a gate-preceded one is refused (a new campaign-validate e2e). Argmax defaults stay byte-identical. closes #215 |
||
|
Brummel
|
abbfbdc208 |
fix(cli): refuse --trace on sweep/walkforward instead of silently dropping it (#168)
The clap help promised "persists each member's/OOS window's taps" and the chart NotFound hint recommended `aura sweep --trace`, but every blueprint-mode arm discarded the flag (run_blueprint_sweep's `let _ = persist`, the --real sugar's persist_taps: vec![]) — the surface advertised trace-recording the code dropped. Refuse --trace up front on both verbs (exit 2, "--trace is not yet available on <verb>; see #224"), mirroring the existing aura run / aura mc refusals; drop the false promise from the two help lines and the chart NotFound hint. Delivering per-member trace-writing is deferred to #224. Collateral: four synthetic-sweep tests that used --trace as a family-name handle move to --name (same handle, still accepted); the walkforward name/trace test now pins the up-front #224 refusal. Workspace suite green (62 groups / 0 failed), clippy -D warnings clean. refs #168 |
||
|
Brummel
|
53bbcedeec |
test(cli): RED — sweep/walkforward must refuse --trace with a #224 pointer (#168)
The clap help promises "persists each member's taps" and the chart hint recommends `aura sweep --trace`, but every blueprint-mode arm discards the flag (run_blueprint_sweep's `let _ = persist`, the --real sugar's persist_taps: vec![]) — the surface advertises trace-recording the code drops. Pin the decided honest behaviour: sweep/walkforward refuse --trace up front (exit 2, naming the flag and pointing forward to the deferred trace feature #224), mirroring the aura run / mc refusals. refs #168 |
||
|
Brummel
|
f6d80486cd |
fix(cli): gate dissolved verbs on project presence, matching campaign run (#218)
A dissolved verb (sweep/generalize/walkforward/mc --real) run outside a project silently created a content-addressed runs/ store in the cwd and ran to completion — Env::runs_root falls back to a relative ./runs with no project, and the four dispatchers skipped the provenance gate that `campaign run` applies first (campaign_run.rs:322-333). It is now refused up front: one gate in the shared validate_and_register_axes chokepoint, after axis validation and before the first store touch (put_blueprint), via a new AxisRegisterError::NoProject mapped to exit 1 carrying "<verb> needs a project …". Placement is load-bearing: axis validation (an UnknownAxis usage error, exit 2) stays ahead of the project gate, so a malformed --axis still exits 2 regardless of project — a malformed command is malformed in any environment and already writes no store. Four per-verb pins assert the refusal (exit 1, "needs a project", no store); the six exit-2 usage-refusal tests stay green. Workspace suite green (62 groups / 0 failed), clippy -D warnings clean. closes #218 |
||
|
Brummel
|
2162bd6d97 |
test(cli): migrate dissolved-verb e2e tests into the project fixture (#218 prep)
The four dissolved verbs (sweep/generalize/walkforward/mc --real) are about to gain a project gate that refuses outside a project (#218). 25 cli_run.rs tests today run a well-formed verb to completion in a bare cwd and pin its grade / store effect — i.e. they pin the pre-gate behaviour. Relocate them into the built project fixture (fresh_project() over the demo-project cdylib, serialized, store reset per test) so they still exercise the verb path once the gate lands. Grades are byte-identical in-project (the fixture Aura.toml overrides no instrument or data path — DEFAULT_DATA_PATH is absolute), verified against the exact pins; local run 113 passed / 0 failed / 0 skipped with archive data. Two new tests pin that the in-project run stamps manifest.project provenance (the sweep chokepoint and generalize's run_blueprint_member path), a path no cli_run.rs test previously exercised. The built_project/project_lock helpers are duplicated into this binary to keep the change inside cli_run.rs; a shared tests/common extraction is a separate hygiene item. refs #218 |
||
|
Brummel
|
210230596e |
feat(cli): default the stop regime on walkforward/mc/generalize
GREEN side of the #217 executable-spec: --stop-length/--stop-k are now optional on the three dissolved verbs; each missing flag independently defaults to the single-sourced regime constants (length 3, k 2.0) at the argv-boundary extractors — the same regime the campaign member runner resolves for an unbound regime. The generated campaign document still binds the regime explicitly, so the default spelling is content-identical to an explicit --stop-length 3 --stop-k 2.0 (pinned by registry dedup in the headline test). The multi-value stop refusal survives byte-identically; sweep is untouched. Sanctioned contract migrations (decision on the issue, 2026-07-09): walkforward_dissolved_refuses_missing_knobs now pins the surviving multi-value refusal; mc_args_from_refuses_missing_knobs pins the default path. Usage strings name the defaults. Held quality nit (deferred to cycle close, minimal-slice constraint): the None=>default / Some=>parse stop block recurs 4x across the walkforward/mc extractors. closes #217 |
||
|
Brummel
|
af8564d20b |
test: RED executable-spec for #217 — stop-less verb invocation defaults the regime
walkforward_stopless_defaults_the_regime_byte_identically_to_explicit_3_2 pins the decided contract (#217 decision comment, 2026-07-09): a walkforward invocation without --stop-length/--stop-k succeeds, each omitted knob defaulting to the single-sourced regime (length 3, k 2.0), and produces a campaign document content-identical to the explicit --stop-length 3 --stop-k 2.0 spelling (registry dedup: exactly one doc). Currently RED: the verb refuses with the requires-knobs usage error. refs #217 |
||
|
Brummel
|
4acea45519 |
chore(cli,docs): #220 cycle-close audit — doc reconcile + dispatch dedup
Cycle-close audit for the verb-axis-generalization cycle ( |
||
|
Brummel
|
b849bb76b9 |
test(cli): weld-gone proof + new refusals for the generic verbs (#220 slice 3)
Task 5 of the verb-axis-generalization plan — the cycle's point, pinned: - walkforward_campaign_runs_an_arbitrary_blueprint and mc_campaign_runs_an_arbitrary_blueprint drive r_breakout_open.json (wrapped channel_hi/channel_lo axes) over the real archive; generalize_campaign_runs_an_arbitrary_blueprint grades r_meanrev_open.json across two instruments. All three invocations were structurally impossible before this cycle (exit 2). - New refusals: campaign mode without --axis (wf/mc), mc --real + --seeds mix, raw-form axis name before data access, unknown axis as a named Knob fault — each mirroring sweep's shipped contract. - Final acceptance gates ran green in the loop: full workspace suite, clippy -D warnings, cargo doc warning-free, zero too_many_arguments allows in verb_sugar.rs, no RGrid in the workspace, no r_sma_open embed at the three verb dispatch sites (the aura graph default embed survives by design). closes #220 closes #214 |
||
|
Brummel
|
e7efe1f765 |
feat(cli): walkforward + mc go blueprint-generic (#220 slice 2)
Tasks 3-4 of the verb-axis-generalization plan: - walkforward: the welded r-sma campaign branch moves out of the None arm into the blueprint mode, split on --real (campaign sugar) vs the unchanged synthetic family path; WalkforwardCmd drops --strategy/--fast/--slow; the full-window clip block, wf_ms_sizes and WINNER_SELECTION_METRIC move unchanged; the --select plateau #215 refusal survives. walkforward_summary_json_from_reports is de-welded: the axis-name list is passed in from the invocation instead of the const AXES r-sma quadruple. - mc: McCmd drops the welded flags and gains --axis; the blueprint arm splits on --real (campaign pipeline: requires --axis, refuses --seeds) vs the unchanged synthetic seed family (--seeds); block-len/ resamples/seed defaults and the --name/--trace refusal stay byte-identical. - translate_walkforward/run_walkforward_sugar and translate_mc/ run_mc_sugar adopt SugarInvocation (+ WfWindows/McKnobs); ZERO #[allow(clippy::too_many_arguments)] remain in verb_sugar.rs (#214 complete). - Tests: wf/mc e2e groups migrate argv-only (grade anchors keep all pinned floats verbatim); collateral usage/retired-token tests assert clap's structural rejection of the deleted flags; the synthetic family control groups pass unmodified. Necessary deviation from the plan's literal step text: re-added #[allow(clippy::type_complexity)] on walkforward_args_from (clippy -D warnings flags the 6-tuple; mirrors its siblings). Held quality nits (validate/strip block now quadruplicated across the four dispatch fns; inline exit-mapping recurrence) are deliberate plan-holds — dedup follows as the rule-of-three cleanup at cycle end. Verification: full workspace suite green in the loop's independent verify; build + clippy -D warnings clean. refs #220, refs #214 |
||
|
Brummel
|
a2294c7c3a |
feat(cli): SugarInvocation + generalize goes blueprint-generic (#220 slice 1)
Tasks 1-2 of the verb-axis-generalization plan (spec auto-signed on a grounding-check PASS, decision log on #220): - verb_sugar gains the shared SugarInvocation<'_> carrier (#214) with VolStop and the factored doc_axes_from/risk_from/probe_params_from helpers; translate_sweep/run_sweep_sugar adopt it (no behaviour change — sweep argv and family output byte-identical), and translate_generalize/run_generalize_sugar follow, dropping their too_many_arguments allows (4 remain for wf/mc). - generalize is de-welded: GeneralizeCmd takes a blueprint positional plus repeatable --axis <wrapped-name>=<value> (one value per axis); --strategy/--fast/--slow are deleted; dispatch_generalize runs the sweep sequence (read, wrapped-namespace validation, canonicalize + topology_hash registration, wrapped_to_raw_axis strip) with a new usage closure. RGrid is deleted workspace-wide. - cli_run generalize tests migrate argv-only; the exact-grade anchor keeps all pinned floats verbatim; generalize_refuses_a_non_r_sma_ strategy is deleted (its premise — a strategy gate — dissolves by design); the multi-value refusal now fires per --axis. Verification: full workspace suite green in the loop's independent verify; build + clippy clean. Held quality finding (kept as a plan-hold): dispatch_generalize duplicates dispatch_sweep's validation/strip block verbatim — dedup follows once wf/mc land the same block (rule-of-three, end of this cycle). refs #220, refs #214 |
||
|
Brummel
|
07a73fb66e |
docs(cli,ledger): sweep #159 demo-retirement drift — stale comments + ledger notes
Cycle-close audit tidy for the #159 demo-retirement arc: behaviour-free reword of prose/comments/docs left pointing at symbols the arc deleted. Adversarially verified by the audit workflow; cargo build --workspace --all-targets green. - README: drop the removed `--harness` token; restrict the legacy built-in form claim to walkforward/mc/generalize (run/sweep no longer resolve one). - glossary: the built-in `--strategy` sweep surface is retired (#159), not a live inline path. - INDEX.md ledger: append [HISTORY] supersession markers (house style) to the dated realization notes advertising the retired `--harness` / built-in `--strategy` CLI forms (cost-flag #153, cycles 0065/0066/0067, name-res, #210 status); historical text preserved. - aura-engine test-fixture docs: drop dead cross-crate refs to the deleted aura-cli fns sample_harness / build_sample and the deleted `macd` composite. - cli_run.rs: reword the #159-cut-2 pin doc to the current clap generic-usage reality (no strategy_from / Strategy::RBreakout); r_sma_sweep_family -> blueprint_sweep_family. - verb_sugar.rs: fix the pre-existing `E[R]` unresolved-doc-link warning. Remaining tidy — the main.rs comment cluster plus its rustdoc link at :1546 (the second cargo-doc warning) — held for a follow-up issue. The dead cost-graph branch is tracked in #221. refs #159 |
||
|
Brummel
|
ee7001647c |
feat(cli): retire the PIP demo pair (sma/momentum) to the generic path (#159 cut 4)
The two legacy pip-scored demos — the SMA-cross (`sma`/`Strategy::SmaCross`) and momentum (`momentum`/`Strategy::Momentum`) — and the whole built-in *synthetic* run/sweep/walkforward/mc + `macd` surface they anchored are dropped, not migrated: both terminated in a `SimBroker(pip_size)` construction-arg node absent from the zero-arg std vocabulary and reported the legacy pip yardstick C10 retired with the realistic broker. With the R-family already migrated to data (cuts 1-3) the `Strategy` enum held only these two, so the retirement collapses the entire built-in-strategy machinery: `Strategy`, `HarnessKind`, `strategy_from`, `run_sweep`/`sweep_family`/`momentum_*`, `run_sample*`/`sample_*`/`sma_cross`, the built-in `walkforward_family`/`sweep_over`/`run_oos` and `run_mc`/`mc_family`, the `macd` composite cluster, and `run_dispatch`/`run_args_from`/`RunCmd.--harness`. The generic surfaces survive untouched: `aura run <bp.json>`, `aura sweep <bp.json> --axis`, `aura walkforward|mc --strategy r-sma --real`, `aura campaign`, `aura generalize`. Grammar consequence (derived forks logged on #159): `run`/`sweep` now require an explicit blueprint (no synthetic default); `sma`/`momentum` join the retired r-breakout/r-meanrev tokens in the generic `Usage:` error, pinned by a new negative (`sweep_and_walkforward_reject_retired_pip_tokens_as_unrecognized`); bare `aura graph` renders the shipped `examples/r_sma_open.json` (loaded via `blueprint_from_json`), and the render self-containment test with it; the `sample-model.json` viewer corpus + its two `.mjs` tests stay frozen as static test data. `member_key`/`MAX_KEY`/`fnv1a64` were removed too. The plan wrongly listed `member_key` as a survivor; after the built-in sweep/mc retire it has no production caller (campaign_run derives its own content key), so it and its unit tests are genuinely dead. The spec-compliance gate surfaced this plan-deviation; it was ratified after verifying the tree stays green. Verified: `cargo build --workspace`, `cargo test --workspace` (0 failed), and `cargo clippy --workspace --all-targets -- -D warnings` all clean. refs #159 |
||
|
Brummel
|
257ab0b9f2 |
feat(std,cli): add the Scale node, retire the r-meanrev demo to data (#159 cut 3)
Cut 3 of the hard-wired demo retirement. Unlike r-sma/r-breakout, r-meanrev could
not round-trip as data: its band computes band_k*sigma (a constant times a stream)
and the closed 22-node std_vocabulary had no way to express it (EqConst is i64->bool,
Bias is clamp(signal/scale,±1), Mul is binary, LinComb is excluded as a
construction-arg node; no Div/Const/constant-emitter). Standard operators are missing
by chance, not by design, so this adds the one r-meanrev needs.
Scale node (aura-std): out = input * factor — one f64 input, one f64 `factor` param,
stateless, warm-up-filtered (the EqConst/Bias shape). A zero-input `Const` emitter was
rejected: the per-cycle eval loop gates every node through an input-firing test
(harness.rs `fires()`), so a node with no input never fires without an engine-core
change; `Scale` fits the existing model and, by IEEE-754 commutativity, reproduces the
retiring LinComb's band byte-for-byte. Rostered (`"Scale" => Scale`); the two
vocabulary count-pins bump 22 -> 23.
r-meanrev migration: r_meanrev_signal(window, band_k) is carved out of the fused
r_meanrev_graph as a #[cfg(test)] price->bias Composite whose band is `Scale` in place
of `LinComb(1)`; production loads the shipped JSON (examples/r_meanrev{,_open}.json),
never a builder. Before the fused builder was deleted, an equivalence test proved the
carved Scale-band signal reproduces its grade byte-for-byte on the synthetic stream
(window=3, band_k=2). Durable survivors: the byte-identity of the examples to the
carve, the loaded-grades-identically-to-carve proof, the closed/open introspect
anchors (mean_window.length, var_window.length, band.factor — band_k is now a
first-class sweepable param), and — re-pointed onto the carve rather than dropped —
the fades-short-above/long-below behavioural test, which becomes the durable
carve-correctness gate the deleted equivalence test used to be.
Deletions + the dead-code cascade the retirement opened: r_meanrev_graph,
r_meanrev_sweep_family, Strategy::RMeanRev (+ all arms). r-meanrev was the last reader
of run_sweep's grid, so the whole vestigial built-in-sweep grid apparatus retires with
it: run_sweep's `grid` param, the sweep call-site grid build, and SweepCmd's
--fast/--slow/--stop-length/--stop-k/--window/--band-k (sma/momentum sweeps build their
own blueprints and ignored these; the fast/slow/stop-* had been vestigial since r-sma's
cut 1b). RGrid the struct SURVIVES — the dissolved `generalize` verb resolves its
candidate from its fast/slow/stop_length/stop_k via generalize_args_from — so only its
window/band_k fields go; its stale doc-comment (r-sma-sweep / persist_traces_r / CLI
flags, all now gone) is rewritten to its true generalize-only role. Also retired as
transitively dead: persist_traces_r and the metrics_object test helper (their only
callers were the deleted families/tests); Add/Gt/Latch/Mul/Sqrt imports are now
#[cfg(test)] (their last production caller was r_meanrev_graph).
Test surface: the two built-in --strategy r-meanrev sweep tests drop; a new negative
pins that --strategy r-meanrev on both sweep and walkforward now falls into the generic
usage error; the grid-flag stray-positional negative is re-pointed to a surviving flag.
Verification (own, not the workflow's report): full `cargo test --workspace` green
(0 failed across 62 result lines); `cargo clippy --workspace --all-targets -- -D
warnings` clean, no dead-code residue; the r-sma + r-breakout anchors and the
dissolved-verb real goldens (generalize still reads RGrid) stay green. Bundled as one
commit (the Scale node's count-pin and the r-meanrev anchors share graph_construct.rs,
so a clean two-commit split was not path-separable); the implement-loop ran all three
plan tasks in one pass and the tree was verified by hand.
refs #159
|
||
|
Brummel
|
5856cadadd |
refactor(cli): retire the r-breakout demo — topology now lives only as data (#159 cut 2)
Cut 2 of the hard-wired demo retirement, applying the r-sma template one level
down. The r-breakout signal leg was fused into r_breakout_graph (which also built
the whole pip/R harness inline), so this cut first carves the signal leg out as a
pure price->bias Composite, ships it as a proven blueprint example, then deletes
the fused builder and the built-in --strategy r-breakout surface. The r-breakout
topology now exists only as data (crates/aura-cli/examples/r_breakout{,_open}.json).
Carved: r_breakout_signal(channel) -> Composite, a verbatim lift of the fused
builder's signal leg (Delay/RollingMax/RollingMin/Gt/Latch/Sub) exposing the raw
exposure as "bias"; the generic wrap_r/run_signal_r path (unchanged) supplies the
pip/R harness. It is #[cfg(test)] — its only role is regenerating the examples and
pinning them to a faithful serialisation; production loads the shipped JSON, never
this builder (so no dead-code residue, no production topology constructor — #159's
invariant).
Proof: before the fused builder was deleted, an equivalence test proved the carved
signal, wrapped and run, reproduces the retiring r_breakout_graph grade byte-for-byte
on the synthetic stream at channel=3 with the shared default R-stop (metric
computation is identical on both paths). That gate is deleted with the builder it
references; the durable anchors that survive are the byte-identity of the shipped
examples to the carve, the loaded-example-grades-identically-to-the-carve proof, and
the closed-ness / open-axis introspect anchors (channel_hi.length, channel_lo.length
per-node — ganging is the separate #61).
Deleted: r_breakout_graph, r_breakout_sweep_family (aura sweep --strategy r-breakout),
Strategy::RBreakout + all four match arms, the r-breakout token from the --strategy
usage strings and docs, and the now-dead --channel flag + RGrid.channel field (its
sole reader was r_breakout_sweep_family). The data successors: aura run
examples/r_breakout.json, and aura sweep examples/r_breakout_open.json --axis
channel_hi.length --axis channel_lo.length.
Test surface: the three built-in-surface sweep tests drop; the walkforward
valid-but-unsupported-token test is re-pointed to momentum (the property is generic
and stays live until Cut 4); and a new negative pins that --strategy r-breakout on
both sweep and walkforward now falls into the generic usage error, not a
special-cased message — catching a half-finished retirement that left the token
parseable.
Verification (own, not the workflow's report): full `cargo test --workspace` green
(0 failed across every binary); `cargo clippy --workspace --all-targets -- -D warnings`
clean, no dead-code residue; the r-sma introspect + grade anchors and the
dissolved-verb real goldens stay green. The implement-loop ran both plan tasks in one
pass (task_range was not honoured); the tree was verified by hand — grep-confirmed the
retired surface is gone and the carve is #[cfg(test)]-scoped.
refs #159
|
||
|
Brummel
|
f7c809e5ce |
refactor(cli): retire the r-sma demo builder — topology now lives only as data (#159 cut 1b)
Cut 1b of the hard-wired demo retirement — the removal half. With the proven
example shipped in cut 1, delete the r-sma builder and its whole transitive
closure so no production function constructs the r-sma topology any more; it
exists only as data (crates/aura-cli/examples/r_sma{,_open}.json).
Deleted: sma_signal (the builder), r_sma_graph, run_r_sma (aura run --harness
r-sma), r_sma_sweep_family (aura sweep --strategy r-sma), the synthetic inline
r-sma walkforward/mc machinery (r_sma_space/r_sma_sweep_over/run_oos_r/
run_mc_r_bootstrap/mc_r_bootstrap_report + the walkforward_family RSma arm), both
fixture emitters, the round-trip + cut-1 byte-fidelity tests, the R_SMA_FAST/
SLOW/BIAS_SCALE consts, and the Strategy::RSma / HarnessKind::RSma / McArgs::RealR
variants. The three dissolved-verb synth sites (generalize/walkforward/mc
--strategy r-sma --real) now embed the shipped open example via include_str!
instead of building it — byte-safe, guarded by the unchanged dissolved-verb
real-archive goldens. The two now-redundant fixtures move to the examples
(deleted, ~40 readers repointed). The --cost-* flags on `aura run` retire with
the harness (their sole consumer was run_r_sma; cost-model coverage survives
engine/std/composites-level).
The retired demo surfaces (--harness r-sma, sweep/walkforward/mc --strategy r-sma
inline) have working data successors: aura run examples/r_sma.json, aura sweep
examples/r_sma_open.json --axis …, and the --strategy r-sma --real campaign path.
Deviation from the plan, noted: StopArm collapsed to StopRule in wrap_r's
signature (its only other variant, Open, died with r_sma_graph) — cleaner than
the planned vestigial single-variant enum; behaviour-preserving, touching
campaign_run.rs and the r-breakout/r-meanrev call sites.
Follow-up (not this cut): wrap_r's cost-node-building branch is now unreachable
(nothing passes Some(CostConfig) after run_r_sma's deletion) yet clippy-clean —
a latent cleanup for when the inline cost surface fully retires.
The stage1-r param-suffix remapping (r_sma_friendly_name + the FAST/SLOW/
STOP_LENGTH/STOP_K_SUFFIX consts) went dead and was removed, closing #167 by
deletion.
Verification (own, not the workflow's report): full `cargo test --workspace`
green; `cargo clippy --workspace --all-targets -- -D warnings` clean, no
dead-code residue; the cut-1 grade anchor (shipped_r_sma_example_reproduces_the_
builtin_grade) and the dissolved-verb real goldens (generalize/walkforward/mc
_dissolves_through_the_campaign_path) stay byte-identical. A first workflow
attempt was discarded after a subagent's `git checkout -- main.rs` recovery
silently reverted earlier tasks' edits (filed on the plugin tracker as
Brummel/Skills#23); this landing is the hardened re-run, verified by hand.
closes #167
refs #159
|
||
|
Brummel
|
d7fa4b4743 |
feat(cli): ship r-sma demo as a proven blueprint-data example (#159 cut 1)
Cut 1 of the hard-wired demo retirement — the copy+prove phase. The r-sma demo strategy now ships as runnable blueprint data under crates/aura-cli/examples/ (r_sma.json closed 2/4, r_sma_open.json open), emitted from the current builder so it is byte-for-byte the builder's serialization. Two load-bearing proofs, captured while the builder is still alive: - byte-fidelity: the shipped examples equal blueprint_to_json(&sma_signal(..)) — the "the copy is identical to the original" evidence the retirement method requires before any original is removed; - grade anchor: `aura run examples/r_sma.json` reproduces the built-in `--harness r-sma` metrics byte-for-byte (expectancy_r/sqn/total_pips/n_trades). Plus two property tests on the shipped copies (beyond the plan, kept as on-topic proof strengthening): the closed example introspects to zero unbound params (genuinely closed), and the open example lists exactly its two SMA-length axes from the public gallery location. Purely additive: the builder (sma_signal) and every inline r-sma surface are retained. The recon showed deleting the builder has a large transitive closure (the synthetic inline surface, the Strategy::RSma enum, ~40 integration tests); that removal is cut 1b, a separate cycle, made safe by the byte-fidelity proof frozen here. Full workspace suite green; clippy -D warnings clean. refs #159 |
||
|
Brummel
|
7f5f576f41 |
feat(cli): dissolve aura mc --strategy r-sma --real (#210)
The FOURTH and LAST verb dissolution. `aura mc --strategy r-sma --real` now runs as thin sugar over the one campaign executor: a generated, content-addressed [std::sweep(argmax), std::walk_forward, std::monte_carlo] campaign document whose terminal monte_carlo stage produces the same StageBootstrap::PooledOos(RBootstrap). The sugar reads it from run.outcome.record.cells[].stages[].bootstrap and reprints the existing mc_r_bootstrap_json — one line, byte-identical to the inline path. Pure sugar, NO executor change (aura-campaign/aura-engine/aura-research diff-free). Mirrors the shipped walkforward dissolution: - verb_sugar.rs: GeneratedMc, translate_mc, register_generated_mc, run_mc_sugar (record-read + one grade line, no member lines). - main.rs: mc_args_from + the dispatch_mc built-in-arm split gated on `--strategy r-sma && --real`. The one new load-bearing decision is the seed mapping: translate_mc sets campaign.seed = the mc --seed. The wf winners are argmax hence deflation-seed-independent (proven by the shipped walkforward anchor), so the pooled OOS series is unchanged while the terminal r_bootstrap at that seed reproduces the inline bootstrap. A new e2e (mc_dissolved_seed_changes_the_bootstrap_draw_not_the_pooled_series) pins exactly this: two --seed values yield identical n_trades but a moved E[R] mean. Fork C retention (as walkforward): the arm splits, run_mc_r_bootstrap / run_mc stay inline for the fenced synthetic/non-real paths (until #159). The three synthetic-r-sma tests use no --real, so they never enter the sugar branch and run the inline path unchanged. --name/--trace are rejected on the sugar path, matching the inline mc r-path. Verification: the committed byte-identity anchor mc_r_bootstrap_real_e2e_pins_the_exact_current_grade runs the real GER40 archive and passes byte-for-byte through the new path; full workspace suite green; clippy --all-targets -D warnings clean. refs #210 |
||
|
Brummel
|
3fc491a058 |
test(cli): pin the exact mc R-bootstrap grade — byte-identity anchor (#210)
The fourth and last verb dissolution (mc's R-bootstrap path) reroutes the inline `walkforward_family -> pooled_oos_trade_rs -> r_bootstrap` through the campaign `[std::sweep(argmax), std::walk_forward, std::monte_carlo]` process, whose terminal monte_carlo stage does the identical `StageBootstrap::PooledOos(r_bootstrap(...))` seeded from the campaign seed. Before that rewire, this pins the EXACT current bootstrap grade of a fixed 2025 GER40 invocation (block_len=5, resamples=1000, seed=42, multi-point fast 3,5 x slow 12,20 grid) so the dissolution must reproduce it byte-for-byte (the acceptance gate). The load-bearing subtlety the dissolution rests on: the campaign path uses ONE campaign seed for both the wf-stage deflation and the mc-stage bootstrap, whereas the inline path uses a fixed DEFLATION_SEED for the wf winners and the mc `--seed` for the bootstrap. This reconciles because the wf winners are argmax (deflation only annotates provenance), so they are seed-independent — the walkforward anchor already proved it (the campaign path at seed 0 reproduced the inline DEFLATION_SEED winners byte-for-byte). So `campaign.seed = mc --seed` leaves the pooled OOS series unchanged and makes the bootstrap match. n_trades=20681 matches the walkforward multi-grid anchor, confirming mc pools the same OOS-R series. Deterministic; gated on the local GER40 archive, skips cleanly on a data refusal. refs #210 |
||
|
Brummel
|
a047af65fb |
feat(cli): dissolve aura walkforward --strategy r-sma --real (#210 T3-T4)
The dispatch split lands: `dispatch_walkforward`'s built-in arm now routes
`--strategy r-sma --real` through the one campaign path (`run_walkforward_sugar`
-> generated campaign doc -> executor -> summary reprint), reproducing the
`{"walkforward":{…}}` line byte-for-byte. Everything else — synthetic (any
strategy) and non-r-sma real — stays on the inline `run_walkforward`, fenced
until #159 (Reading A). The committed anchor
`walkforward_real_e2e_pins_the_exact_current_grade` passes: the dissolved path
reproduces windows=9, stitched_total_pips, oos_r, and the param_stability means
exactly. `--select plateau:*` is refused (exit 2 -> #215, Fork B); a multi-value
stop is refused (Fork A).
Two plan guesses were wrong and the anchor + real-data e2e caught both; the fixes
are documented inline:
- Window resolution: the plan mirrored generalize's `(--from,--to) -> (f,t)`
shortcut, but the inline `wf_full_span` ALWAYS clips `--from`/`--to` to the
archive's actual first/last bar (`probe_window`), even when both flags are
explicit — a holiday/weekend edge shifts every IS/OOS window's calendar
placement (stitched_total_pips diverged by ~5.16M, param_stability means intact,
confirming only window boundaries moved). The dissolved path now calls
`source.full_window(env)` unconditionally, matching the inline behaviour;
verified byte-for-byte against the still-fenced inline path.
- Reducer axis match: the campaign path records the blueprint axes wrapped
(`sma_signal.fast.length`) while the stop rides unwrapped via the risk regime, so
the reducer's exact-name match panicked on real data. Fixed with the existing
`campaign_run::raw_matches_wrapped` (#203) wrap-convention helper.
Tests: the dissolution-proof e2e `walkforward_dissolves_through_the_campaign_path`
(one process + one campaign + one campaign-run auto-registered, windows=9) plus
front-end refusal coverage (multi-value stop, `--select plateau`, missing knobs,
empty `--real`, unknown `--select`, `--name`/`--trace` exclusivity). Full workspace
suite green; `clippy --all-targets -D warnings` clean.
refs #210
|
||
|
Brummel
|
96dc7832e5 |
test(cli): pin the exact walkforward grade — byte-identity anchor (#210)
The third verb dissolution (walkforward) reroutes the inline per-window IS-refit (`walkforward_family` -> `select_winner` -> OOS run -> stitch/pool) through the campaign `std::walk_forward` stage. Two summary fields have no campaign equivalent today: `stitched_total_pips` (the runner discards the raw OOS pip curve, `oos_equity: vec![]`) and pooled `oos_r` (only a following `std::monte_carlo` stage produces it). Before the executor work that closes those gaps, this pins the EXACT current grade of a fixed 2025 GER40 invocation so the dissolution must reproduce it byte-for-byte (the acceptance gate). The grid is deliberately multi-point (fast 3,5 x slow 12,20): it makes the per-window winner selection non-degenerate (`param_stability` varies across the 9 windows), so a winner-pick divergence between the inline and campaign refit cannot ship green — the same gap the generalize anchor closed for its sibling shape-only test. Deterministic (C1); gated on the local GER40 archive, skips cleanly on a data refusal. refs #210 |
||
|
Brummel
|
f5bb38ed59 |
test(cli): dissolution-proof e2e for generalize (#210 T4)
`generalize_dissolves_through_the_campaign_path` asserts the verb now runs through the one campaign path: exactly one generated process document, one generated campaign document (carrying the --name handle and the stop as a non-empty risk regime), and one campaign-run record are durably auto-registered, plus the campaign-path family set (1 CrossInstrument grade family + 2 per-instrument Sweep families). The observable proof the inline path is gone. refs #210 |
||
|
Brummel
|
256ec7320c |
feat(cli): dissolve aura generalize into the campaign path (#210 T3)
`dispatch_generalize` synthesizes the bare `sma_signal` blueprint, resolves the
window (explicit --from/--to, else the first symbol's full window), and calls
the new `run_generalize_sugar`; the inline `run_generalize` is deleted. The
sugar builds the selection-bearing campaign doc, runs it through
`run_campaign_returning`, reprints the byte-identical `{"generalize":…}` +
`{"family_id":…}` lines, and persists the CrossInstrument family.
The exact-grade anchor stays green through the path shift — proving the two stop
mechanisms (former grid axis, now the risk-regime seam -> StopRule::Vol) yield
identical R. The family pin relaxes to "exactly one CrossInstrument family"
since the campaign executor now also persists a per-instrument Sweep family per
cell (the durable audit trail; #210 Q4). Two new e2e pin the window fallback and
the family member attribution order.
Deviations from the plan (necessary, verified): cross_instrument_members returns
Vec<aura_engine::RunReport> (aura_registry::RunReport is not pub); a shared
validate_before_register helper is factored out of run_sweep_sugar (DRY); the
run_generalize doc-comment and the now-unused `generalization` import removed.
refs #210
|
||
|
Brummel
|
f3f32b8269 |
test(cli): pin the exact generalize grade — byte-identity anchor (#210)
The current `aura generalize` binds its protective stop as a grid axis (`r_sma_sweep_family` with `stop_open=true`); the pending generalize dissolution rebinds the same stop through the risk-regime seam (`RiskRegime::Vol -> StopRule::Vol`). The existing generalize e2e (`generalize_grades_a_candidate_across_two_instruments`) asserts shape only — `worst_case` and the per-instrument R floats are never checked — so a divergence between the two stop bindings would ship green. This adds `generalize_real_e2e_pins_the_exact_current_grade`, capturing the exact current R grade of the identical invocation (GER40 0.0105637…, USDJPY 0.0057959…, worst_case 0.0057959…, sign_agreement 2). Green today against the inline path; after the dissolution the same command must reproduce these bytes — the acceptance gate that catches a stop-mechanism divergence loudly instead of silently. Resolves the generalize-dissolution grounding block (the unpinned stop-mechanism R equivalence) by the risk-regime characterization-pin pattern. refs #210 |
||
|
Brummel
|
c62f7e9878 |
feat(cli): bind the resolved regime stop and stamp it into the manifest (#210 T4)
The runtime-binding slice — the regime's stop now actually reaches the graph and
the record:
- `StopArm { Open, Bound(StopRule) }` replaces `wrap_r`'s `stop_open: bool` so a
bound stop carries its params; all call sites threaded (the r-sma sweep path
stays Open, every bound site passes the resolved rule).
- `run_blueprint_member` takes a `stop: StopRule` and stamps `stop_length` /
`stop_k` into the member manifest (the ratified C18 extension — the stop is no
longer invisible on the campaign/blueprint path); its callers thread the
default, and `CliMemberRunner` maps `cell.regime` (None -> baked default,
Some(Vol) -> that vol stop).
- The characterization pin flips from asserting the stop's ABSENCE to requiring
the stamped default; a real two-regime e2e pins that each regime stamps its
own k.
Known deferred gap (concern, documented descope): `persist_campaign_traces`
re-runs the nominee to persist its traces and — since `CellRealization` carries
no regime — binds that re-run to the DEFAULT stop, so a non-default-regime
nominee's persisted trace would reflect the wrong stop. Only the persist path
(non-empty `persist_taps`) is affected; the core run/stamp/generalize path is
correct. Filed as a follow-up. Full workspace suite green.
refs #210
|
||
|
Brummel
|
455916db73 |
test(cli): pin the stop-less campaign member manifest (risk-regime grounding)
The dissolved real-data sweep member manifest carries only the swept signal axes today — no stop param; the R-path stop is a wrap_r constant baked outside the signal param_space. The existing member-line pin (sweep_real_blueprint_member_lines_pin_the_dissolved_contract) locates each binding by `find`, so it never pinned this absence. This characterization pin does, as the grounding ratification for the risk-regime-axis design pass: the resolved-regime manifest stamp will flip it to require the stop key — the ratified additive contract extension, exactly like the instrument stamp. refs #210 |
||
|
Brummel
|
c9d962f0ea |
fix(cli): sweep-entry ergonomics — honest refusals, no store litter (#210 c0110 fieldtest)
Three test-first fixes from the cycle-0110 fieldtest findings: - Axis-name preflight on the dissolved sweep arm: every --axis name is checked against the wrapped probe namespace (the --list-axes names) BEFORE the #203 strip; an unknown or raw-form name is refused house-style echoing exactly what the user typed, with a --list-axes pointer — no more stripped-then-mangled fragments from the deep referential refusal. Data-free, fires before the archive is touched. - Validate-before-register in the sweep sugar: intrinsic doc checks + executor preflight + the pure bind_axes coverage check (reused, not re-inlined; now pub(crate)) all run in-memory before register_generated — a refused invocation leaves no generated document in the content-addressed store and no campaign_runs.jsonl line (previously P2/P3 refusals littered runs/campaigns, one doc malformed). - blueprint_slot_prose: the sweep/walkforward/mc loaded-blueprint slots share one dispatch-boundary presenter — an op-script array gets a targeted 'run aura graph build first' hint, every load error goes through blueprint_load_prose + the missing-project hint, and the raw {e:?} Debug leak (#184's pattern, deliberately left on these three slots back then) is gone. Two stale Debug-leak pins flipped, two new refusal tests. Suite 1060/0, clippy -D warnings clean. refs #210 |
||
|
Brummel
|
b7aaa0ba59 |
feat(cli): dissolve the real-data blueprint sweep into the campaign path (#210 T4)
aura sweep <bp.json> --real ... --axis ... now runs as thin sugar over its generated campaign document through the one campaign executor: the dispatch registers the canonical blueprint (topology_hash IS content_id_of over the same canonical bytes, so the strategy ref resolves against the existing single store write), strips the CLI's wrapped axis names to the raw campaign namespace (#203 bind convention), converts the probed archive window back through the ingest seam's own epoch_ns_to_unix_ms (made pub; C3 — the seam owns the ms/ns convention, never an inline division), and hands off to verb_sugar::run_sweep_sugar. The inline real-data arm is gone; the synthetic arm and --list-axes are untouched. campaign_run gains RunPresentation {Full, MemberLinesOnly}: run_campaign keeps Full (record line unchanged, project gate unchanged); the sugar path runs MemberLinesOnly — the generated doc's emit already limits stdout to family_table member lines, the mode suppresses only the final record line. The record append is identical in both modes; no project gate on the by-id path (the dissolved verb works project-less, exactly as before). The characterization pin survives the re-cut with exactly the one sanctioned assertion flip (instrument-absence -> instrument==GER40) and now also pins the generated artifacts: one process + one campaign document auto-registered (dedup on a second identical invocation), raw axis keys in the stored document, the stored window a ms sub-range of the request, C1 at the member-report level with the registry's run-suffix increment isolated. Suite 1055/0, clippy clean, gated real-data paths executed locally. refs #210 |
||
|
Brummel
|
8235cee4d7 |
test(cli): characterization pins ahead of the sweep dissolution
Two green pins the sweep-dissolution cycle rests on (refs #210): - sweep_real_blueprint_member_lines_pin_the_inline_contract: the inline real-data blueprint sweep prints one member line per grid point in odometer axis order, carries the swept bindings in manifest.params, persists exactly one Sweep family — and never stamps an instrument key (RunManifest::instrument stays None on this path). - campaign_run_real_e2e_sweep_gate_walkforward (extended): every family_table member line emitted by the real campaign member-runner path is stamped with the instrument under test. Together they pin the one behavioural delta between the two paths (the additive instrument stamp) that the dissolution's parity contract is phrased against. |
||
|
Brummel
|
61c8828fad |
feat(cli): house-style loader-error prose + missing-project hint on the run path
aura run leaked the raw LoadError Debug form (UnknownNodeType("..."))
while graph build already phrased the same fault house-style (#162
convention: engine error types are Display-free, the CLI presents them).
Route the run path through the shared blueprint_load_prose presenter (now
pub(crate), env-agnostic) so both surfaces phrase a LoadError identically
and neither leaks the Debug token (#184). Add unresolved_namespace_hint:
when the unresolved id looks project-namespaced (::) and no project was
loaded (env.provenance() is None), append a hint pointing at the missing
Aura.toml — the newcomer running the host from the wrong directory now
gets an orienting diagnosis instead of a bare type error (#185). The hint
is applied on both the run and graph-build paths. RED-first: the two
existing run-path pins were flipped to assert the prose + absence of the
Debug token before implementing; suite green (1043/0).
closes #184
closes #185
|
||
|
Brummel
|
9402afe958 |
polish(cli): normalize usage-message casing to the clap house style
Every hand-rolled usage line now reads 'Usage: aura <verb> ...' — capital prefix plus the program name, the shape clap itself and the built-in run branch already spoke. 10 literal edits: 6 lowercase 'usage:' flips (graph introspect, run blueprint x2, runs family, mc blueprint, mc built-in), 4 bare closures gain the prefix (sweep/walkforward, blueprint + built-in branches); the mc built-in line names the program in both |-separated alternatives. Refusal diagnostics, the conforming run_args_from literal, clap output, and the fieldtests corpus stay byte-identical; exit codes keep the usage=2 / runtime=1 partition. Lockstep pin: the single casing-sensitive assertion (mc --real) flips to capital 'Usage'. Three additive regression tests pin the house style itself (bare-closure prefix, both mc alternatives, mc blueprint branch) — the pre-existing pins checked only verb tokens / a single Usage occurrence; deviation from the spec's no-new-test-bytes wording recorded on the issue. Verification: cargo test --workspace rc=0 (cli_run 133 passed), clippy -D warnings clean, grep gates exact (zero lowercase 'usage:' literals; 11 'Usage: aura' source lines), smoke 'aura sweep --strategy bogus' prints the house-style line and exits 2. closes #179 |
||
|
Brummel
|
4de6d5cbad |
rename: retire the stage1-* family for the r-family (r-sma / r-breakout / r-meanrev)
The stage1 ordinal has been dead vocabulary since the C10 reframe dropped the two-stage research model: a 1 structurally implies a 2 that no longer exists. The family is renamed by its live discriminator - the R yardstick - with members named by their signal, uniform with their signal-named siblings (sma, macd, momentum): - selectors: stage1-r -> r-sma, stage1-breakout -> r-breakout, stage1-meanrev -> r-meanrev (old tokens are usage errors, exit 2 - no silent alias) - identifiers: Strategy::RSma/RBreakout/RMeanRev, HarnessKind::RSma, r_sma_*/r_breakout_*/r_meanrev_*, wrap_r, run_signal_r, RGrid, R_SMA_* - persisted identity: the sma_signal composite (param prefix sma_signal.fast.length / .slow.length; fixtures regenerated; content-ids shift - no test pins a literal hash, the registry parses no record names) - e2e test files git-mv'd to r_sma_e2e / r_breakout_e2e / r_meanrev_e2e - dead Stage-1/Stage-2 prose reworded to post-C10 vocabulary across rustdoc, Cargo.tomls, ledger live lines, and the glossary (historical entries stay; fieldtests corpus untouched) - CLAUDE.md invariant 7 rewritten to record the ratified C10 reframe faithfully (the token-swap alone would have laundered the retired gated-currency/realistic-broker design into unmarked live prose); the unbacked account-mode clause dropped Verification: cargo build/test --workspace green (51 targets), clippy -D warnings clean, doc build clean, acceptance grep gate leaves exactly the one resampling-stage false positive (harness.rs), smoke: --harness r-sma runs, --harness stage1-r exits 2 with the new usage line. Decision log: forks and rationale recorded on the issue (reconciliation + implementation-phase comments). closes #174 |
||
|
Brummel
|
88e9c9a539 |
fix(cli): symmetric sweep/mc family-verb polish (closed error, mc seed label, mandatory-knob docs)
Three polish items surfaced by the World/C21 milestone fieldtest:
- A sweep of a fully-bound (closed) blueprint is refused up front with a clear
"this blueprint is fully bound; nothing to sweep" message — the symmetric
inverse of blueprint_mc_family's closed-blueprint guard — instead of leaking a
misleading UnknownKnob for a knob that is not unknown but bound out.
blueprint_sweep_family's error contract widens to String; the sweep terminal's
BindError (a genuine unknown/kind-mismatch axis) still surfaces verbatim.
- mc reproduce lines now carry each member's seed=<N> label. Monte-Carlo members
hold no tuning params, so the member label was blank; the seed is each draw's
realization identity. Sweep / walk-forward labels (which echo params) unchanged.
- docs (design ledger + glossary): state that every open knob --list-axes prints
is mandatory on sweep / walkforward — a subset is refused with the missing knob
named (BindError::MissingKnob), there is no default; pin a knob you do not want
to vary with a single-value axis.
RED-first for the two behaviour items. One existing unknown-axis E2E is re-pointed
from the closed fixture (which now hits the new "nothing to sweep" pre-check) to
the open fixture; it still exercises UnknownKnob("nope") specifically — resolve
raises UnknownKnob in phase 1 before MissingKnob in phase 2 — so the property is
preserved, not weakened.
closes #178
|
||
|
Brummel
|
e20dd9ae49 |
Merge main into the clap-adoption cycle — re-integrate the two parallel CLI fixes
main advanced with two `fix(cli)` commits ( |
||
|
Brummel
|
fa42bf3878 |
feat(0099): CLI exit-code split (iteration 2) — usage=2, runtime=1
Apply the clean exit-code partition (#175 deviation #8) on top of the clap migration: runtime failures now exit 1, usage errors stay exit 2, with no same-class inconsistency. Partition (attribution principle): exit 2 = a command-line fault (a bad flag value, an inapplicable combination, or the content of an argv-named file); exit 1 = the command was well-formed but the environment / recorded state it needs is missing, or piped stdin data is bad. The 6 boundary cases the spec's "parse-time vs run-time" rule under-specified are resolved on #175 (malformed/open argv blueprint -> usage 2; stdin op-script content -> runtime 1; applicability refusals -> usage 2; unknown --axis / --metric -> usage 2). Flipped 2->1 (runtime): no local data, no recorded geometry, no recorded run/family, "run has no tap named", trace-name collision, family/trace persist-write failures (normalizing the lone generalize inconsistency -- all persist -> 1 together), missing/corrupt content-addressed store state (the reproduce path), chart-read failures, and graph stdin op-script content + stdin-read I/O. reproduce-diverged stays exit 1. Usage sites (clap parse, argv-applicability guards, the dual-grammar blueprint-file read/parse) stay exit 2. Tests: ~8 domain-refusal pins flipped Some(2)->Some(1) (message substrings unchanged), the four runtime tests renamed _exit_2 -> _exit_1, and a partition property test added. Orchestrator fix after the loop: 3 sibling no-data skip-guards (run_real_with_trace + the two generalize cross-instrument tests) that the plan's Task-1 list missed -- left at Some(2), they would fall through to assert Some(0) and FAIL on a data-less machine; flipped to Some(1) for consistency with the flipped code. Deferred (cosmetic, filed forward): full "Usage:" / "usage:" / bare error-message casing normalization -- no functional impact, high pin-churn. Verified green (orchestrator, this session): cargo build --workspace, cargo clippy --workspace --all-targets -- -D warnings, and cargo test --workspace all clean (0 failed across every test binary). This completes the #175 clap-adoption cycle (iteration 1 = clap migration at 366170a; iteration 2 = this exit-code split). refs #175 |
||
|
Brummel
|
366170aeab |
feat(0098): CLI clap migration (iteration 1) — scoped help, --version, GNU flags
Replace the hand-rolled aura-cli argv parser with a clap derive parser. Delivers, all from one declarative source: - scoped `aura <sub> --help` (stdout, exit 0) — each subcommand's own Options section; retires the #131 uniform-global-blob help - `--version` / `-V` → `aura 0.1.0` (stdout, exit 0) - GNU `--flag=value`, the `--` end-of-options terminator, and long-option abbreviation (`--harn` → `--harness`) Architecture: one root Cli/Command enum, one `*Cmd` struct per subcommand. The four dual-grammar subcommands (run/sweep/walkforward/mc) map the loaded-blueprint branch as an optional `[blueprint]` positional + a post-parse `is_file()` dispatch on a single-sourced predicate; each built-in handler re-asserts a stray-positional guard (refuse-don't-guess) so clap's optional positional cannot silently swallow a typo (pinned by two new E2E tests). The execution layer (run_*/emit_*/runs_*/reproduce_*) is untouched — only arg-plumbing changes, via thin `*_from` adapters reusing the existing value helpers. Exit codes behaviour-preserved: clap parse errors exit 2 (matching the old usage-error=2); domain refusals stay exit 2; reproduce-diverged stays exit 1. The exit-code split (2=usage / 1=runtime) is iteration 2. clap is admitted under the C16 per-case review (research-side CLI, invariant 8 untouched — a dev-loop compile tax, not a frozen-artifact tax). Forced plan corrections (implementer; verified against the diff): - Step 6's deletion of the `parse_*_args` fns orphaned ~34 grammar unit tests that called them; those were deleted and that grammar coverage moved to the renegotiated cli_run.rs E2E pins (`parse_select` kept + retested). - run_malformed_cost_value_...: a bad cost value is now a clap value-parse error (no "Usage:" line); renegotiated to pin exit 2 + empty stdout + the flag name "--cost-per-trade"; the units note stays pinned via `run --help`. - `allow_hyphen_values` on the three cost flags so `--cost-per-trade -0.5` reaches the non-negativity guard (else clap rejects -0.5 as an unknown option). - Deleted the now-dead `RealWindowGrammar` (clippy -D warnings). Orchestrator additions after the loop: - Enabled long-option abbreviation (`infer_long_args` on the root — one attribute, it propagates to subcommands) + a test, closing spec acceptance criterion 4 which the plan's RED tests did not cover. Decided to deliver rather than amend the spec: the two ratified decisions conflict on this low-tier item — Fork B (GNU compliance, getopt_long abbreviates) vs F5 (LLM/automation-first, human-only niceties deprioritized) — and the cost tiebreaker (one attribute, not per-struct work) resolves to deliver. Carry-on debt (not this iteration): the error-message "Usage:"/"usage:"/bare casing is now mixed (clap's capitalized "Usage:" beside preserved lowercase aura messages); a deliberate quality-hold since the casings are pinned by renegotiated/preserved tests. Cosmetic; a candidate for the iteration-2 exit-code-split cleanup. Verified green (orchestrator, this session): cargo build --workspace, cargo clippy --workspace --all-targets -- -D warnings, and cargo test --workspace all clean (0 failed across every test binary, including the renegotiated cli_run pins and the new --version / scoped-help / GNU-flag / dual-grammar tests). refs #175 |
||
|
Brummel
|
6a775f700f |
fix(cli): walkforward validates its --axis grid once, not once per window
`aura walkforward <bp> --axis <name>=<v>` emitted a rejected axis's diagnostic more than once (a racy 1-3x) when the grid did not resolve. `blueprint_walkforward_family` resolved the axes *inside* the per-window closure, and `walk_forward` fans that closure out across the windows in parallel — so several windows each `eprintln!` the same `BindError` and `exit(2)` before any one exit tears the process down. Axis resolution is window-independent, so it is hoisted to a single dispatch-boundary pre-flight (mirroring `aura sweep`, which validates its axes once before any member runs): the first in-sample window is resolved up front, surfacing the `BindError` exactly once; the per-window closure then resolves already-validated axes and cannot re-raise (`expect`). RED-first: cli_run.rs `aura_walkforward_emits_an_unknown_axis_rejection_once` drives the verb 20x and asserts the rejection is single every run (the emit count was racy, so a single invocation would be a flaky assertion). Surfaced by the World/C21 milestone fieldtest. closes #177 refs #170 |
||
|
Brummel
|
dec07809df |
fix(cli): aura run refuses an open blueprint with a clean exit 2, not a panic
`aura run <blueprint.json>` bootstraps the loaded signal over the EMPTY param point (a closed-blueprint assumption) and `.expect()`s `compile_with_params`, which panics (exit 101, ParamArity) when the blueprint carries >=1 free knob. The sibling `aura mc` already refuses an open blueprint cleanly at the dispatch boundary; `run` now does the same. The `["run", ..]` `.json`-discriminator arm probes the wrapped param_space (`blueprint_axis_probe`, the single source of that wrap) and, when it is non-empty, emits a "run requires a closed blueprint … N free knob(s) — bind them or use `aura sweep --axis`" diagnostic and exits 2 — mirroring `blueprint_mc_family`'s closed-guard. No panic on a naive consumer's first move (author an open blueprint, run it). RED-first: cli_run.rs `aura_run_rejects_an_open_blueprint_without_panicking` pins exit 2 + the named requirement + no `panicked`. Surfaced by the World/C21 milestone fieldtest. closes #176 refs #170 |
||
|
Brummel
|
ff20f74e8f |
feat(0097): IS-refit walk-forward over a loaded blueprint — aura walkforward <bp.json> --axis
Ship the World/C21 milestone's last piece (Arm A, #173): a true in-sample-refit walk-forward over a loaded blueprint. `aura walkforward <blueprint.json> --axis <name>=<csv> [--axis …] [--select argmax|plateau:mean|plateau:worst]` rolls 24/12/12 IS/OOS windows over the synthetic span; per IS window it re-optimizes the blueprint's params over the user's --axis grid (the #169 prefixed names), selects the winner by sqn_normalized, runs it out-of-sample, and aggregates the per-window OOS reports into a FamilyKind::WalkForward family — content-addressed (put_blueprint) and aura-reproduce bit-identical. Structure mirrors the hard-wired stage1-r walk-forward arm one substitution deep: the generic engine walk_forward driver + select_winner + the synthetic schedule are reused; only the per-window IS-sweep (blueprint_sweep_over) and OOS-run (run_oos_blueprint) source the loaded blueprint via cycle-0096's blueprint_axis_probe. reproduce_family_in gains a WalkForward arm reconstructing each OOS window's windowed slice from the member manifest's window bounds (winner params via the shared manifest->cells recovery, as the MonteCarlo/Sweep arms do). Panic-safety: blueprint_sweep_over returns Result<_, BindError>; a bad/unknown --axis is a clean in-closure exit 2 (mirroring the hard-wired arm's select_winner handling), never a panic — pinned by the rejects_an_unknown_axis E2E. A no-axis invocation (nothing to re-fit) is a usage error; a malformed blueprint is rejected at the dispatch boundary (exit 2, no panic). Reduce-mode members carry R-metrics (no raw pip curve) → empty stitched equity; oos_r is the meaningful summary (C10 R-first yardstick). Held quality finding adjudicated: run_oos_blueprint's plan-prescribed redundant `pip: f64` (== data.pip_size()) removed — derived internally to match its siblings run_oos_r / blueprint_sweep_over, which also drops the too_many_arguments band-aid #[allow] (8->7 args). Two plan transcription typos the loop caught and fixed verified correct: BindError renders via Debug ({e:?}), parse_select via map_err (Result, not Option). Verified myself: cargo build --workspace, cargo clippy --workspace --all-targets -- -D warnings, cargo test --workspace all green; 2 new unit tests + 5 new cli_run E2Es (persist shape, reproduce bit-identical, no-axis, malformed-safety, unknown-axis) pass; no regression in the hard-wired walkforward / sweep / mc / reproduce suites. Engine/registry untouched (invariants 1/2/8/9). closes #173 |
||
|
Brummel
|
bd2003e217 |
feat(0096): list a loaded blueprint's sweepable axes — aura sweep <bp.json> --list-axes
A data-level author can now discover a loaded blueprint's open sweepable knobs before sweeping: `aura sweep <blueprint.json> --list-axes` prints one `<name>:<kind>` line per open knob, in param_space() order, then exits 0. Before this the only way to learn the names was to guess a `--axis` and read the MissingKnob/UnknownKnob error — which names only the rejected knob, never the valid set. Prerequisite for #173 (IS-refit walk-forward must know which axes to re-fit per in-sample window). Design (derived, recorded on #169): the query lands on the SWEEP verb, not `graph introspect`, because the accepted axis names are produced by the sweep's own harness-wrapping (wrap_stage1r nests the loaded signal as a BlueprintNode::Composite named "stage1_signal", so collect_params prefixes them — stage1_signal.fast.length, not the raw fast.length). Only the sweep verb owns that wrapping. A new blueprint_axis_probe helper single-sources the wrapped probe and now subsumes the two previously-inline copies (blueprint_sweep_family + blueprint_mc_family), so listed == swept by construction (and stays so across #159's harness retirement) — no second source of truth. --list-axes is a standalone query (valueless flag handled before the per-flag value-consume; rejected with exit 2 if combined with any other flag). A closed blueprint prints nothing (exit 0). A malformed blueprint is rejected at the existing dispatch boundary (exit 2, UnknownNodeType), never a panic. Tests: unit blueprint_axis_probe (prefixed names + empty for closed) + unit parse grammar (standalone), and five cli_run E2Es — prints names, closed-empty, rejects-other-flags, malformed-safety (no panic), and a round-trip that DERIVES the names from --list-axes and feeds them back into a working sweep (pins the listed==swept invariant #169 exists to close). Verified myself: cargo build --workspace, cargo test --workspace, cargo clippy --workspace --all-targets -D warnings all green; no regression in the refactored sweep/mc probes. Engine and registry untouched (invariants 1/8/9). closes #169 |
||
|
Brummel
|
14e27a978b |
feat(0095): monte-carlo over a loaded blueprint — aura mc <bp.json> --seeds N
The MC half of #170 (World/C21): `aura mc <closed blueprint.json> --seeds N` builds a FamilyKind::MonteCarlo family from a loaded CLOSED blueprint. Each seed draws a distinct synthetic price walk (synthetic_walk_sources, the mc_family pattern), so members differ in the DATA (not just a manifest label) — the sim is deterministic (C1), so MC variation must come from re-drawing the input. Draws run disjointly in parallel via the shared monte_carlo helper (invariant 1). The canonical blueprint is stored once keyed by the shared topology_hash (the 0094 hook), and reproduce_family_in gains a MonteCarlo realization branch that reconstructs each member's seed-driven walk from manifest.seed — so a persisted mc family 'aura reproduce's bit-identically (the store-hook proof). Closed-blueprint only: MC binds no axis, so a free knob has no binder; an open blueprint yields a named Err (rendered to stderr + exit 2 at the run_blueprint_mc boundary, mirroring blueprint_sweep_family — no hidden process exit in the pure builder), pre-empting the compile_with_params arity panic. Two quality-gate-driven improvements over the ratified plan, both applied and verified: (1) delegate the disjoint draws to the engine monte_carlo seam instead of a hand-rolled loop (my plan's 'closure-shape mismatch' rationale was wrong — the Fn(u64,&[Scalar]) closure ignores the unused base_point, exactly as mc_family does); (2) the closed-blueprint guard returns Result rather than exiting inside the builder, matching the sweep sibling's contract and making the rejection unit-testable. Engine + registry untouched (invariants 8/9). Verified: cargo build --workspace clean; cargo clippy --workspace --all-targets -- -D warnings clean; cargo test --workspace green (5 new tests: the parser, anti-degenerate family, MC reproduce round-trip, and two E2Es incl. the open rejection; no regressions). refs #170 |
||
|
Brummel
|
cec7811c32 |
audit(0094): cycle close — content-addressed reproduction; ledger refreshed, coverage closed; rm ephemeral spec/plan
Architect drift review (4751d4b..HEAD): drift_found — ledger-currency only; the
code is clean.
What holds (architect, confirmed by diff + green tests):
- Inv 8 / C9: aura-engine byte-untouched; content_id + sha2 in aura-cli, the
store in aura-registry.
- Inv 9: Registry::{put,get}_blueprint is a dumb bytes-by-key store (opaque
String keyed by String hash; no sha2, no parse, no sha256==hash verify — the
caller owns the hash). No domain knowledge crept into the engine or registry.
- C1: the run_blueprint_member extraction is behaviour-preserving (keystone
blueprint_sweep_member_equals_single_run_and_shares_topology_hash unchanged +
green); serde_json/float_roundtrip serves determinism (disk==recomputed), not a
mask (1-ULP canary green).
- C11/C12: one blueprint stored per family, keyed by the shared topology_hash.
Resolutions:
- [ledger] Refreshed C18 + C24: added the cycle-0094 Realization (the
content-addressed store runs/blueprints/<hash>.json + aura reproduce), and
retired the now-false "content-addressed identity / full reproduction deferred"
claims (INDEX.md:1261, C24 deferred block). Only the structural-axis /
whole-harness content-id (and the debug-name-in-id question) genuinely remains,
filed forward (#171, #170).
- [debt->closed] Added the two untested refuse-don't-guess E2E arms
(aura_reproduce_rejects_a_non_generated_family — a hard-wired sweep member has
no topology_hash; aura_reproduce_rejects_a_missing_stored_blueprint — the store
was deleted), and a cross-surface unit pin
(topology_hash_is_the_content_id_of_the_canonical_form) so the two acc-1 paths
cannot silently drift.
- [scaling] MC/walk-forward-from-blueprint (#170) will need the same put_blueprint
hook or reproduce won't cover them — noted on #170.
No regression scripts in this project (Test is the gate): full workspace suite
green (51 suites), clippy -D warnings clean, binary verified end-to-end.
Ephemeral spec/plan git-rm'd per the lifecycle convention; durable rationale is in
the ledger. The World/C21 milestone stays open (#170 remains).
closes #158
|
||
|
Brummel
|
008692c043 |
feat(0094): content-addressed reproduction — aura reproduce <family-id> (iter 1)
Restore C18 reproducibility for generated (blueprint-sweep) runs (#158, acc 2): a sweep's topology is stored once, content-addressed by the topology_hash the manifest already carries, and `aura reproduce <family-id>` re-derives every persisted member bit-identically (C1) from the stored blueprint. - aura-registry: a dumb content-addressed bytes store — Registry::put_blueprint / get_blueprint / blueprints_dir over runs/blueprints/<hash>.json. No sha2 dep (caller owns the hash; reproduction's bit-identical compare is the integrity check); absent id -> Ok(None) (treat-as-empty, like load). - aura-cli: run_blueprint_member extracted from blueprint_sweep_family's member closure (behaviour-preserving; keystone stays green) so the sweep AND reproduce re-run the identical reduce-mode path — bit-identity by construction. The store write hooks the sweep persist seam (one blueprint per family, shared topo). reproduce_family_in loads each member's blueprint by hash, reconstructs the point from the recorded params (point_from_params, inverse of zip_params, over the WRAPPED signal's param_space so the prefixed knob names match), re-runs, and compares metrics. `["reproduce", id]` dispatch arm; refuse-don't-guess on an unknown id / missing stored blueprint (exit 2), DIVERGED -> exit 1. Dependency decision (per-case review, ratified): enabled serde_json `float_roundtrip`. Stored member metrics round-trip through families.jsonl as f64 JSON; the default parser can be 1 ULP off, which would make disk-loaded != recomputed and "bit-identical" (C1) physically impossible. The feature guarantees exact f64 round-trip. Blast-radius clean (full workspace suite green); it serves determinism, the invariant reproduction rests on. Canary: f64_blueprint_param_survives_store_round_trip_bit_identically. Two plan bugs caught + fixed by the implement loop and verified: reproduce's param_space must come from the wrapped signal (prefixed names, the #167 lockstep), and the family id is 0-indexed (live-0 / smacross-0, not -1). Verified: full workspace suite green (51 suites); clippy -D warnings clean; and the binary end-to-end — `aura sweep` (3 members, one shared stored blueprint) then `aura reproduce live-0` (3/3 bit-identical incl the open-at-end member), unknown id -> exit 2. acc 1 (graph introspect --content-id) + acc 3 (Tier-1 id stability) land in iteration 2 (plan 0094b). refs #158 |
||
|
Brummel
|
481add0ca2 |
feat(0093): sweep-from-blueprint — aura sweep <blueprint.json> over named axes
Cycle 2 of the World/C21 milestone (#166): the World now orchestrates FAMILIES from blueprint-data. `aura sweep <blueprint.json> --axis <name>=<csv> …` loads an OPEN signal blueprint, grids the by-name axes against its param_space, builds each member via cycle-1's wrap_stage1r seam, and aggregates to a FamilyKind::Sweep family — byte-identical in shape to the hard-wired sweep. - `parse_sweep_blueprint_args` + `SweepBlueprintArgs`: repeatable `--axis name=csv` (lexed i64-then-f64; resolve_axes kind-checks), `--name`/`--trace`, the shared `--real/--from/--to` accumulator → DataChoice. - `blueprint_sweep_family` (pure) + `run_blueprint_sweep` (IO) mirror stage1_r_sweep_family with three deviations: by-name axes (verbatim param_space names, e.g. `stage1_signal.fast.length`), reload-per-member (Composite is !Clone — its build closures are Box<dyn Fn>, #164 — so a fresh signal is re-loaded from the doc per member), and a per-member shared topology_hash. Members share the hash (same signal topology, only params vary); member_key distinguishes them. - The `aura sweep <file.json>` `.json`-discriminator arm mirrors cycle 1; the hard-wired `--strategy` sweep is untouched (#159). Task-3 design gap resolved: a sweep needs an OPEN blueprint (open knobs to grid), distinct from cycle-1's fully-bound run fixture (a bound blueprint has an empty param_space). New `tests/fixtures/stage1_signal_open.json` = blueprint_to_json(stage1_signal(None,None)); a sweep must bind ALL open knobs, so the E2E grids fast×slow (2×2 = 4 members). Verified end-to-end by running the binary: `aura sweep stage1_signal_open.json --axis stage1_signal.fast.length=2,4 --axis stage1_signal.slow.length=8,16 --trace f` persists a discoverable 4-member Sweep family; an unknown axis fails clean. Member equivalence vs run_signal_stage1r + shared-hash/distinct-member_key unit-tested. Full workspace suite green; clippy --all-targets -D warnings clean. refs #166 |
||
|
Brummel
|
2b1c24668d |
test(0092): run-from-blueprint E2E + bit-identical keystone + demo fixtures
Closes the cycle-1 test coverage for #165. The `aura run <blueprint.json>` CLI arm itself (the .json dispatch + parse_blueprint_run_args + BlueprintRunArgs) shipped in ff4a1b3; this adds the coverage that proves it and the in-repo demo blueprint: - crates/aura-cli/tests/fixtures/stage1_signal.json — the canonical SMA-cross signal (blueprint_to_json of stage1_signal(2,4)); emit_demo_signal_fixture (#[ignore]) regenerates it. - crates/aura-cli/tests/fixtures/unknown_node.json — one type set to "Nope" for the fail-clean guard. - loaded_signal_runs_bit_identical_to_rust_built (the keystone, C1): a signal serialized -> loaded via blueprint_from_json -> run is bit-identical to its Rust-built twin, incl. topology_hash (64-hex). - topology_hash_is_stable_and_distinguishes (#158 anchor property). - cli_run.rs E2E: `aura run <demo>` exits 0 with topology_hash in the manifest; an unknown-node blueprint fails clean with UnknownNodeType (the #160 closed-set guard at the data-plane face, invariant 9). Verified end-to-end by actually running the binary: `aura run stage1_signal.json` -> RunReport with a 64-hex topology_hash, sim-optimal+risk-executor broker, R metrics; `aura run unknown_node.json` -> UnknownNodeType("Nope"), exit 2. Full workspace suite green; clippy --all-targets -D warnings clean. refs #165 |
||
|
Brummel
|
d925a747cb |
feat(0086): cost-flag CLI ergonomics — units note, named diagnostics, non-R-harness guard
#153 (friction surfaced by the milestone-#148 close fieldtest). Three cost-flag fixes on `aura run`, all on the pure parse_run_args grammar: - Units discoverable: a COST_FLAGS_NOTE appended to the run usage + --help (price units, charged in R as cost/|entry-stop|, per-ENGINE-cycle carry, >= 0, stage1-r only). - A negative rate names the cause + flag ("<flag> must be non-negative, got <v>") via a factored parse_nonneg_rate helper, not a bare usage dump. - A cost flag on a non-R harness (sma/macd) is refused (exit 2, "cost flags require an R-evaluator harness (stage1-r)") instead of silently ignored — the named spec_gap, decided reject/fail-loud (refuse-don't-guess, recorded on #153), pinned by a new C10 ledger note. Internal RunArgs/RunData/HarnessKind gain #[derive(Debug)] (behaviour-neutral) so the pure parse_run_args unit tests can use expect_err. Verified independently: cargo test -p aura-cli green, the eight stage1-r cost goldens byte-unchanged; clippy -p aura-cli --all-targets -D warnings clean. Only the grammar/usage strings and the previously-unguarded sma-cost path changed; the cost-model graph and R math are untouched. closes #153 |
||
|
Brummel
|
f5e00a9c72 |
feat(0085): per-cycle-held accrual — CarryCost + ChargeMode (approach B)
Cycle 5 of milestone #148 (cost-model graph in R): the per-cycle-held accrual mechanism — C10's "per-cycle-held factors accrue over the hold". A carry/funding cost is charged for every cycle a position is held, so the net_r_equity equity curve bleeds continuously over the hold (approach B, "ja, mach B") rather than stepping at close. Mechanism (logged on #148): ChargeMode { AtClose, PerHeldCycle } is a property of the cost factor, read by the one shared CostRunner (no second runner type). The AtClose arm is the pre-cycle-5 eval tokens VERBATIM (IEEE-754 byte-identity). The PerHeldCycle arm accrues `per` into `acc` each held cycle, dumps the accrued total into `cum` at close (resetting acc), and marks the open position via a growing open_cost_in_r. The bleed lives in open_cost_in_r, which the net_r_equity tap already subtracts — so summarize_r and the CLI net_eq wiring are UNCHANGED, and the cycle-0083/0084 net_expectancy_r goldens (-614.3134020253314 flat, -615.0304388396047 composed) + the C18 no-cost golden stay byte-identical. CarryCost is a ConstantCost twin differing only in charge_mode() -> PerHeldCycle (a labelled stress parameter, the flat base of the accrual family). Wired through the existing cost_graph/CostSum aggregation; a per-trade ConstantCost and a per-held-cycle CarryCost compose in one run (the costs sum per-field — the composed golden is exactly the sum of the two single-cost charges). New run-path --carry-per-cycle flag (sweep/walkforward/mc pass None, as the existing cost flags do). Tests: 2 RED-first unit B-proofs (open_cost grows over the hold, dumps the total at close, acc resets between trades — the discriminator a scalar net_expectancy_r cannot see); the CarryCost twin's 5 unit tests; 2 captured net_expectancy_r goldens (-768.2095026600887 carry, -1383.7939051991182 composed); a net_r_equity-bleeds-over- the-hold integration test (the terminal open hold's r_equity-net_r_equity gap strictly grows); plus negative-rate-refused-exit-2, monotone-in-rate, and zero-rate-exactly-free guards. Verified: cargo test --workspace (0 failures), cargo build --workspace, cargo clippy --workspace --all-targets -- -D warnings clean. summarize_r / aura-analysis untouched. Deferred to later #148 cycles (logged there): notional-based carry, the calendar-aware overnight swap, sweep-path cost. refs #148 |