feat(strategies): convert rapid, bdd, discuss to vBRIEF-centric outputs (#363, #365, #366)

[MScottAdams]

Summary

Convert rapid, bdd, and discuss strategies to vBRIEF-centric outputs, eliminating hand-authored markdown artifacts in favor of structured vBRIEF JSON files.

Closes #363, Closes #365, Closes #366
Parent: #338

Changes

rapid.md (#363)

• Step 3: write �brief/specification.vbrief.json (Light path, draft status) instead of hand-authoring SPECIFICATION.md
• Run ask spec:render to produce SPECIFICATION.md as a read-only export
• Output artifacts: specification.vbrief.json is the primary artifact

bdd.md (#365)

• Step 4: write locked decisions to �brief/proposed/{feature}-bdd.vbrief.json with Scenarios and LockedDecisions narratives
• Steps 1/2: direct test files to project test directory, not specs/{feature}/acceptance-tests/
• Eliminated {feature}-bdd-context.md as a hand-authored artifact
• specs/ folder is not created by BDD

discuss.md (#366)

• Output: write locked decisions to �brief/proposed/{scope}-context.vbrief.json with LockedDecisions narrative
• Eliminated {scope}-context.md as a hand-authored artifact
• Promoted SHOULD rule to MUST for vBRIEF narrative persistence
• Updated chaining gate artifact registration to reference vBRIEF path

Tests

• 12 new content tests in ests/content/test_strategy_outputs.py

Checklist

• CHANGELOG.md updated under [Unreleased]
• ask check passes (1426 passed, 2 xfailed)
• Conventional commit message
• Feature branch (not master/main)
• Only touches: strategies/rapid.md, strategies/bdd.md, strategies/discuss.md, tests/content/test_strategy_outputs.py, CHANGELOG.md

[greptile-apps]

Greptile Summary

Converts rapid.md, bdd.md, and discuss.md to vBRIEF-centric outputs, replacing hand-authored markdown artifacts (SPECIFICATION.md, {feature}-bdd-context.md, {scope}-context.md) with structured vbrief/proposed/*.vbrief.json files and rendered exports. The strategy changes are internally consistent with interview.md's vBRIEF model, and the 12 new content tests correctly guard the key constraints. All findings are P2 test-quality suggestions.

Confidence Score: 5/5

Safe to merge — all findings are P2 test-quality suggestions with no correctness or data-integrity impact.

All three strategy documents are consistently updated to vBRIEF-centric outputs, cross-references within the files are correct, and the 12 new content tests pass against the current state. The two flagged items are both P2: a broad string assertion in one test and class-level eager file reads that would obscure errors at collection time. Neither blocks merge.

tests/content/test_strategy_outputs.py — minor test quality improvements suggested (assertion specificity and import-time file reads).

Important Files Changed

Filename	Overview
strategies/rapid.md	Step 1 updated to reference vbrief/specification.vbrief.json; Step 3 now writes the vBRIEF first and renders SPECIFICATION.md via task spec:render; Output Artifacts updated correctly.
strategies/bdd.md	Steps 1/2 test storage location moved from specs/ to project standard test dir; Step 4 decisions now written to vbrief/proposed/{feature}-bdd.vbrief.json with Scenarios and LockedDecisions narratives; Output Artifacts and Anti-Patterns updated consistently.
strategies/discuss.md	Output artifact changed from {scope}-context.md to vbrief/proposed/{scope}-context.vbrief.json; SHOULD rule promoted to MUST for vBRIEF narrative persistence; chaining gate artifact registration path updated.
tests/content/test_strategy_outputs.py	12 new content tests verifying vBRIEF-centric output references; class-level file reads execute at import time (fragile if strategy files are missing); test_scenarios_narrative assertion is too broad.
CHANGELOG.md	Changelog entry added under [Unreleased] per repo convention.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    subgraph rapid ["rapid strategy"]
        R1["Step 1: State Goal\n→ vbrief/specification.vbrief.json"]
        R2["Step 2: Minimal Interview"]
        R3["Step 3: Generate specification.vbrief.json\n(draft status)"]
        R_RENDER["task spec:render\n→ SPECIFICATION.md (read-only export)"]
        R4["Step 4: Build"]
        R5["Step 5: Evaluate"]
        R1 --> R2 --> R3 --> R_RENDER --> R4 --> R5
    end
    subgraph bdd ["bdd strategy"]
        B1["Step 1: Identify Scenarios\n→ tests/ (standard dir)"]
        B2["Step 2: Write Failing Tests\n→ tests/"]
        B3["Step 3: Run Tests — Surface Ambiguity"]
        B4["Step 4: Lock Decisions\n→ vbrief/proposed/{feature}-bdd.vbrief.json\n  Scenarios + LockedDecisions narratives"]
        B5["Step 5: Generate Spec"]
        B6["Step 6: Chain → interview.md"]
        B1 --> B2 --> B3 --> B4 --> B5 --> B6
    end
    subgraph discuss ["discuss strategy"]
        D1["Open / Explore / Challenge Vagueness"]
        D2["Lock Decisions\n→ vbrief/proposed/{scope}-context.vbrief.json\n  LockedDecisions narrative (MUST)"]
        D3["Verify (Feynman check)"]
        D4["Chain → interview.md Chaining Gate"]
        D1 --> D2 --> D3 --> D4
    end
    B6 --> interview["interview.md Chaining Gate"]
    D4 --> interview
    R5 -->|"graduate"| interview
    style R_RENDER fill:#fef08a,stroke:#ca8a04,color:#000
    style B4 fill:#6ee7b7,stroke:#059669,color:#000
    style D2 fill:#6ee7b7,stroke:#059669,color:#000
    style R3 fill:#6ee7b7,stroke:#059669,color:#000

Loading

Prompt To Fix All With AI

This is a comment left during a code review.
Path: tests/content/test_strategy_outputs.py
Line: 102-105

Comment:
**Overly broad assertion — passes on section headings, not the narrative key**

`assert "Scenarios" in self._text` matches the word wherever it appears (e.g. `"### Step 1: Identify User Scenarios"`), so the test would still pass even if the backtick-wrapped `` `Scenarios` `` vBRIEF narrative key was removed. A tighter check on the narrative definition prevents a false-passing regression.

```suggestion
    def test_scenarios_narrative(self) -> None:
        assert "`Scenarios`" in self._text, (
            "strategies/bdd.md must reference Scenarios narrative in vBRIEF"
        )
```

How can I resolve this? If you propose a fix, please make it concise.

---

This is a comment left during a code review.
Path: tests/content/test_strategy_outputs.py
Line: 31-34

Comment:
**Class-level reads fail at collection time, not test time**

`_text = _read("strategies/rapid.md")` (and the equivalent in the other two classes) runs at module import. If any strategy file is missing or renamed, pytest raises a `FileNotFoundError` during collection, reporting a single import error rather than three individual test failures — making it harder to tell which file caused the problem.

Consider deferring the read to a `@classmethod` setup or a module-level fixture:

```python
@pytest.fixture(scope="class", autouse=True)
def text(self, request):
    request.cls._text = _read("strategies/rapid.md")
```

Or keep it simple with a lazy property — the key is that a missing file surfaces as a named test failure rather than an opaque collection error.

How can I resolve this? If you propose a fix, please make it concise.

_{Reviews (3): Last reviewed commit: "fix: address Greptile review findings (b..." | Re-trigger Greptile}

[MScottAdams]

Rebase-only force-push — no logic changes, rebased onto updated phase2/vbrief-cutover after PR #376 merge.