refactor: schema-driven pred create (11K → 2.3K lines) (#1008)

* docs: add schema-driven create.rs refactor design spec

Design for replacing the 11K-line create.rs with a schema-driven
generic dispatch using existing registry factory functions. Targets
~73% line reduction by eliminating the 5,400-line match statement.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: update design spec with Codex review findings

- Fix problem count (177 → 189)
- Expand type parser table from 16 to 52 types in 3 categories
- Add serde edge cases section (try_from, from, skip, custom BigUint)
- Add Appendix A with complete flag→field mismatch list
- Add generic resolution notes for W, W::Sum, Vec<Vec<W>>

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Phase 1: align create CLI flags with schema names

* Phase 4: add CreateArgs flag map

* Phase 2: add schema field parser registry

* Phase 3: add schema-driven create builder

* Phase 6: wire schema-driven create with fallback

* Phase 5: generate create help from schema examples

* refactor: expand schema-driven create coverage

* refactor: remove legacy create match arms

* chore: clean schema-driven create refactor

* docs: update skills and CLAUDE.md for schema-driven create

- add-model Step 4.5: replace "add match arm" with schema-driven instructions
- review-structural check 13: replace grep for problem name with schema field→flag verification
- CLAUDE.md Extension Points: document schema-driven create convention

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* chore: dedup all_data_flags_empty, remove legacy flag_map aliases, enforce schema naming

- Remove 16 duplicate field checks in all_data_flags_empty()
- Remove backward-compat aliases ("sets", "query") from flag_map();
  schema-derived names ("subsets", "query-attribute") are canonical
- Map schema field "sets" → "subsets" in help_flag_name for consistency
- Add CLAUDE.md rule: CLI flag names must match schema field names

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: GiggleLiu

.claude/CLAUDE.md

@@ -172,9 +172,11 @@ Reduction graph nodes use variant key-value pairs from `Problem::variant()`:
 
 ### Extension Points
 - New models register dynamic load/serialize/brute-force dispatch through `declare_variants!` in the model file, not by adding manual match arms in the CLI
+- **CLI creation is schema-driven:** `pred create` automatically maps `ProblemSchemaEntry` fields to CLI flags via `snake_case → kebab-case` convention. New models need only: (1) matching CLI flags in `CreateArgs` + `flag_map()`, and (2) type parser support in `parse_field_value()` if using a new field type. No match arm in `create.rs` is needed.
+- **CLI flag names must match schema field names.** The canonical name for a CLI flag is the schema field name in kebab-case (e.g., schema field `universe_size` → `--universe-size`, field `subsets` → `--subsets`). Old aliases (e.g., `--universe`, `--sets`) may exist as clap `alias` for backward compatibility at the clap level, but `flag_map()`, help text, error messages, and documentation must use the schema-derived name. Do not add new backward-compat aliases; if a field is renamed in the schema, update the CLI flag name to match.
 - Aggregate-only models are first-class in `declare_variants!`; aggregate-only reduction edges still need manual `ReductionEntry` wiring because `#[reduction]` only registers witness/config reductions today
 - Exact registry dispatch lives in `src/registry/`; alias resolution and partial/default variant resolution live in `problemreductions-cli/src/problem_name.rs`
-- `pred create` UX lives in `problemreductions-cli/src/commands/create.rs`
+- `pred create` schema-driven dispatch lives in `problemreductions-cli/src/commands/create.rs` (`create_schema_driven()`)
 - Canonical paper and CLI examples live in `src/example_db/model_builders.rs` and `src/example_db/rule_builders.rs`
 
 ## Conventions

.claude/skills/add-model/SKILL.md

@@ -175,18 +175,15 @@ The CLI now loads, serializes, and brute-force solves problems through the core
 
 ## Step 4.5: Add CLI creation support
 
-Update `problemreductions-cli/src/commands/create.rs` so `pred create <ProblemName>` works:
+CLI creation is **schema-driven** — `pred create <ProblemName>` automatically maps `ProblemSchemaEntry` fields to CLI flags via `snake_case → kebab-case` convention. No match arm in `create.rs` is needed.
 
-1. **Add a match arm** in the `create()` function's main `match canonical.as_str()` block. Parse CLI flags and construct the problem:
-   - Graph-based problems with vertex weights: add to the `"MaximumIndependentSet" | ... | "MaximalIS"` arm
-   - Problems with unique fields: add a new arm that parses the required flags and calls the constructor
-   - See existing arms for patterns (e.g., `"BinPacking"` for simple fields, `"MaximumSetPacking"` for set-based)
+1. **Ensure CLI flags exist** in `problemreductions-cli/src/cli.rs` (`CreateArgs` struct) for each field in your `ProblemSchemaEntry`. The flag name must match the field name via `snake_case → kebab-case` (e.g., field `edge_weights` → flag `--edge-weights`). If a flag already exists with the right name, you're done.
 
-2. **Add CLI flags** in `problemreductions-cli/src/cli.rs` (`CreateArgs` struct) if the problem needs flags not already present. Update `all_data_flags_empty()` accordingly.
+2. **Add new CLI flags** only if the problem needs flags not already present. Add them to `CreateArgs` and update `all_data_flags_empty()` accordingly. Also add entries to the `flag_map()` method on `CreateArgs`.
 
-3. **Update help text** in `CreateArgs`'s `after_help` — add the new problem to the "Flags by problem type" table in `problemreductions-cli/src/cli.rs` (search for `Flags by problem type`).
+3. **Add type parser support** if the field uses a type not yet handled by `parse_field_value()` in `create.rs`. Check the existing type dispatch table — most standard types (`Vec<i32>`, `Vec<usize>`, `Vec<(usize, usize)>`, graph types, etc.) are already covered. Only add a new parser for genuinely new types.
 
-4. **Schema alignment**: The `ProblemSchemaEntry` fields should list **constructor parameters** (what the user provides), not internal derived fields. For example, if `m` and `n` are derived from a matrix, only list `matrix` and `k` in the schema.
+4. **Schema alignment**: The `ProblemSchemaEntry` fields should list **constructor parameters** (what the user provides), not internal derived fields. For example, if `m` and `n` are derived from a matrix, only list `matrix` and `k` in the schema. Field names must match the struct field names exactly (used for JSON serialization and CLI flag mapping).
 
 ## Step 4.6: Add canonical model example to example_db
 
@@ -315,8 +312,8 @@ Structural and quality review is handled by the `review-pipeline` stage, not her
 | Wrong `declare_variants!` syntax | Entries no longer use `opt` / `sat`; one entry per problem may be marked `default` |
 | Forgetting CLI alias | Must add lowercase entry in `problem_name.rs` `resolve_alias()` |
 | Inventing short aliases | Only use well-established literature abbreviations (MIS, SAT, TSP); do NOT invent new ones |
-| Forgetting CLI create | Must add creation handler in `commands/create.rs` and flags in `cli.rs` |
-| Missing from CLI help table | Must add entry to "Flags by problem type" table in `cli.rs` `after_help` |
+| Forgetting CLI flags | Schema-driven create needs matching CLI flags in `CreateArgs` for each `ProblemSchemaEntry` field (snake_case → kebab-case). Also add to `flag_map()`. |
+| Missing type parser | If the problem uses a new field type, add a handler in `parse_field_value()` in `create.rs` |
 | Schema lists derived fields | Schema should list constructor params, not internal fields (e.g., `matrix, k` not `matrix, m, n, k`) |
 | Missing canonical model example | Add a builder in `src/example_db/model_builders.rs` and keep it aligned with paper/example workflows |
 | Paper example not tested | Must include `test_<name>_paper_example` that verifies the exact instance, solution, and solution count shown in the paper |

.claude/skills/review-structural/SKILL.md

@@ -62,7 +62,7 @@ Only run if review type includes "model". Given: problem name `P`, category `C`,
 | 10 | Re-exported in `models/mod.rs` | `Grep("{P}", "src/models/mod.rs")` |
 | 11 | Variant registration exists | `Grep("declare_variants!|VariantEntry", file)` |
 | 12 | CLI `resolve_alias` entry | `Grep("{P}", "problemreductions-cli/src/problem_name.rs")` |
-| 13 | CLI `create` support | `Grep('"{P}"', "problemreductions-cli/src/commands/create.rs")` |
+| 13 | CLI `create` support | Schema-driven: verify each `ProblemSchemaEntry` field has a matching CLI flag in `CreateArgs` (field `snake_case` → flag `kebab-case`). Check `flag_map()` includes the flag. If the field type is unusual, verify `parse_field_value()` handles it. |
 | 14 | Canonical model example registered | `Grep("{P}", "src/example_db/model_builders.rs")` |
 | 15 | Paper `display-name` entry | `Grep('"{P}"', "docs/paper/reductions.typ")` |
 | 16 | Paper `problem-def` block | `Grep('problem-def.*"{P}"', "docs/paper/reductions.typ")` |