chore: remove deprecated speckit/OMC/beads toolchain artifacts

.claude/RULES_INDEX.md

@@ -123,7 +123,7 @@ Quick version (skills auto-invoke when you mention these):
 - [ ] Link to detailed docs rather than duplicate content
 - [ ] Update skill descriptions with current year (2026) in research queries
 - [ ] Archive rules that haven't been referenced in 60 days
-- [ ] Run `bd stats` monthly to verify rule effectiveness
+- [ ] Review `/sw-audit` findings monthly for rule effectiveness
 
 ---
 

.claude/agents/contract-stability.md

@@ -2,7 +2,7 @@
 name: contract-stability
 description: >
   Specialized agent for floe contract regression testing.
-  Invoked by /speckit.test-review command in parallel with other agents.
+  Invoked by /sw-verify command in parallel with other agents.
   Validates CompiledArtifacts schema stability and cross-package contracts.
 tools: Read, Grep, Glob, Bash
 model: opus

.claude/agents/plugin-quality.md

@@ -2,7 +2,7 @@
 name: plugin-quality
 description: >
   Specialized agent for floe plugin testing completeness.
-  Invoked by /speckit.test-review command in parallel with other agents.
+  Invoked by /sw-verify command in parallel with other agents.
   Validates that all 11 plugin types are tested with proper lifecycle coverage.
 tools: Read, Grep, Glob, Bash
 model: opus

.claude/agents/test-debt-analyzer.md

@@ -2,6 +2,7 @@
 name: test-debt-analyzer
 model: sonnet
 description: Consolidated test debt analysis - flakiness, isolation, edge cases, duplication, coverage gaps. Use for comprehensive test quality audits.
+tools: Read, Glob, Grep, Bash
 ---
 
 # Test Debt Analyzer

.claude/rules/epic-recovery.md

@@ -15,7 +15,7 @@ fi
 1. **Read the state file** to understand current context
 2. **Continue implementing automatically** - do NOT wait for user prompt
 3. **Follow the workflow**:
-   - Sync from Linear: `bd linear sync --pull`
+   - Check workflow state in `.specwright/state/workflow.json`
    - Find next ready task (status: backlog/unstarted)
    - Implement (TDD, SOLID, atomic commits)
    - Update Linear status to Done

.claude/rules/quality-escalation.md

@@ -307,7 +307,7 @@ When escalation identifies a problem that won't be fixed immediately:
 
 1. **Create a GitHub Issue** in the repo with label `tech-debt` or `architecture`
 2. **Add a code comment** with the issue reference: `# TODO(FLO-XXX): description`
-3. **Record in session notes** via `bd update` or notepad
+3. **Record in session notes** via notepad or session memory
 
 NEVER leave a workaround without a tracking issue.
 
@@ -316,7 +316,7 @@ NEVER leave a workaround without a tracking issue.
 ## Enforcement
 
 This rule is enforced by:
-- **Pre-PR review** (`/speckit.test-review`): Checks for assertion weakening
+- **Pre-PR review** (`/sw-verify`): Checks for assertion weakening
 - **Architect verification**: Final gate before completion claims
 - **Critic agent**: Reviews for workaround anti-patterns
 - **Constitution compliance**: PR review checks for principle violations

.claude/rules/skill-invocation.md

@@ -10,12 +10,20 @@
 | `**/assets.py`, `**/resources.py`, `**/io_managers.py` | `dagster-skill` | Orchestration |
 | `**/test_*.py`, `**/conftest.py` | `testing-skill` | Test writing |
 
-## Skill Chains (See `.claude/skill-chains.json`)
+## Specwright Workflow
+
+| Command | Purpose | When |
+|---------|---------|------|
+| `/sw-design` | Research codebase, design solution | New feature or significant change |
+| `/sw-plan` | Break design into work units with specs | After design approval |
+| `/sw-build` | TDD implementation of work unit | Implementation phase |
+| `/sw-verify` | Quality gates (tests, security, wiring, spec) | Pre-PR validation |
+| `/sw-ship` | Create PR with evidence | After all gates pass |
+
+## Skill Chains
 
 | Chain | Skills | Trigger |
 |-------|--------|---------|
-| `epic-planning` | specify→clarify→plan→tasks→taskstolinear | "plan epic" |
-| `pre-pr` | test-review + wiring-check + merge-check (parallel) | "pre-pr check" |
 | `dbt-work` | dbt-skill→pydantic-skill | `*.sql` files |
 | `k8s-deploy` | helm-k8s-skill | `charts/**` |
 | `plugin-dev` | pydantic-skill→dagster-skill→testing-skill | `plugins/**` |
@@ -51,18 +59,6 @@ For less frequent technology work, reference docs are in `docs/reference/`:
 - `docs/reference/duckdb-lakehouse.md` - DuckDB compute
 - `docs/reference/arch-review.md` - Architecture review (use `tech-debt-review --arch`)
 
-## OMC Agent Integration
-
-For generic tasks, use OMC agents instead of custom:
-
-| Task | OMC Agent |
-|------|-----------|
-| Code quality review | `oh-my-claudecode:code-reviewer` |
-| Architecture analysis | `oh-my-claudecode:architect` |
-| Build fixes | `oh-my-claudecode:build-fixer` |
-| Security review | `oh-my-claudecode:security-reviewer` |
-| Codebase search | `oh-my-claudecode:explore` |
-
 ## Custom Agents (floe-Specific)
 
 Keep custom agents for project-specific concerns:

.claude/rules/test-organization.md

@@ -390,7 +390,7 @@ def test_create_catalog():
 
 ## Directory Structure Validation
 
-These checks are enforced by `/speckit.test-review`:
+These checks are enforced by `/sw-verify`:
 
 ### DIR-001: No `__init__.py` in test directories
 

.claude/skills/README.md

@@ -214,7 +214,7 @@ Skills are automatically available. When working on a component:
 | Define semantic layer | cube-semantic-layer | dbt-transformations |
 | Implement compiler | pydantic-schemas | All (integration) |
 | Code quality review | tech-debt-review | - |
-| Pre-PR health check | tech-debt-review | speckit-test-review |
+| Pre-PR health check | tech-debt-review | /sw-verify |
 | Monthly audit | tech-debt-review (--all) | - |
 
 ## Skill Development Guidelines