From smorch-dev
BRD acceptance criteria traceability. Every AC-N.N has a matching @AC-N.N tagged test. Greps BRD vs tests to compute coverage. Caps Engineering Q2 at 6 if coverage <90 percent.
How this skill is triggered — by the user, by Claude, or both
Slash command
/smorch-dev:brd-traceabilityThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
**Purpose:** Every Business Requirements Document acceptance criterion has a tagged test. Nothing ships without traceability.
Purpose: Every Business Requirements Document acceptance criterion has a tagged test. Nothing ships without traceability.
Pillar: Boris #4 — Verification Before Done + EO-Brain integration Time cost: 5-10 minutes per BRD at start of sprint. Zero cost at PR time (just grep).
Every BRD lives at architecture/brd.md (or docs/brd.md). Every user story has numbered acceptance criteria. Every AC has a test tagged @AC-{story}.{criterion}.
## Story 1: Password Reset
**As a** logged-out founder
**I want to** reset my password via email
**So that** I can regain access if I forget it
### Acceptance Criteria
- **AC-1.1** User enters email, receives reset link within 60 seconds
- **AC-1.2** Link expires after 1 hour
- **AC-1.3** Password must be ≥8 chars with 1 number
- **AC-1.4** Success page shows login CTA
// tests/auth/password-reset.test.ts
describe('Password Reset', () => {
it('sends reset email within 60s @AC-1.1', async () => { ... });
it('expires link after 1 hour @AC-1.2', async () => { ... });
it('enforces password strength @AC-1.3', async () => { ... });
it('redirects to login on success @AC-1.4', async () => { ... });
});
Run before every PR that claims to "complete" a BRD story:
# For each AC-N.N in the BRD, find matching test
grep -oE "AC-[0-9]+\.[0-9]+" architecture/brd.md | sort -u > /tmp/brd-acs.txt
grep -orE "@AC-[0-9]+\.[0-9]+" tests/ | grep -oE "AC-[0-9]+\.[0-9]+" | sort -u > /tmp/test-acs.txt
echo "=== ACs in BRD but no test ==="
comm -23 /tmp/brd-acs.txt /tmp/test-acs.txt
echo "=== Tests tagged but no BRD AC ==="
comm -13 /tmp/brd-acs.txt /tmp/test-acs.txt
Decision rules:
| Sprint phase | Minimum coverage |
|---|---|
| Spike / prototype | 50% (happy path only) |
| Feature complete | 90% (happy + error) |
| Pre-ship | 100% (all ACs mapped) |
If architecture/brd.md is missing for the feature being built:
Shipping without a BRD forces Product hat ≤ 4 (see product-hat.md red flags).
EO-Brain Phase 4 (Architecture) produces the BRD. The handover-bridge skill converts EO-Brain output → Claude Code starter:
architecture/brd.md@AC-N.N tags + .skip stubsBRD: 12 ACs
Tests tagged: 12 @AC refs
Coverage: 100% → Engineering hat Q2 = 10
BRD: 12 ACs
Tests tagged: 7 @AC refs
Missing: AC-2.3, AC-2.4, AC-3.1, AC-4.2, AC-5.1
Coverage: 58% → Engineering hat Q2 = 4, PR blocked
BRD: 8 ACs (older)
Tests tagged: 11 @AC refs (newer, includes AC-9.1 that's not in BRD)
Action: Update BRD to include AC-9.1 or remove test
it() block. Otherwise failure granularity is lost.Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub smorchestra-ai/smorch-dev --plugin smorch-dev