F168: onboarding quality taxonomy

.beads-sdp-mapping.jsonl

@@ -311,3 +311,12 @@
 {"sdp_id":"00-167-02","beads_id":"sdplab-mh02","updated_at":"2026-05-04T05:17:09Z"}
 {"sdp_id":"00-167-03","beads_id":"sdplab-809n","updated_at":"2026-05-04T05:17:09Z"}
 {"sdp_id":"00-167-04","beads_id":"sdplab-sfxm","updated_at":"2026-05-04T05:17:09Z"}
+{"sdp_id":"00-168-00","beads_id":"sdplab-o8gk","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-01","beads_id":"sdplab-f16801","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-02","beads_id":"sdplab-f16802","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-03","beads_id":"sdplab-f16803","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-04","beads_id":"sdplab-f16804","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-05","beads_id":"sdplab-f16805","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-06","beads_id":"sdplab-f16806","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-07","beads_id":"sdplab-f16807","updated_at":"2026-05-13T05:45:00Z"}
+{"sdp_id":"00-168-08","beads_id":"sdplab-f16808","updated_at":"2026-05-13T05:45:00Z"}

.cursor/README.md

@@ -18,7 +18,8 @@ logic unless the harness format itself requires it.
 
 ## Usage
 
-Use `@` commands for the main SDP flows:
+Use `@` commands for the main SDP flows. These are **harness commands**, not
+`./.sdp/bin/sdp` CLI calls:
 
 ```text
 @feature "description"
@@ -30,6 +31,29 @@ Use `@` commands for the main SDP flows:
 @operate "deploy or release task"
 ```
 
+If this repo is newly installed, start with:
+
+```bash
+agent -p "@build 00-XXX-YY"
+```
+
+`00-XXX-YY` is a workstream ID placeholder and requires a real `00-XXX-YY` from
+an internal queue.
+
+Cursor is currently treated as a **secondary validator** for SDP dispatch and is
+**untested for primary automation**.
+
+Use it for independent checks and reviews.
+If you need primary dispatch, use Claude Code first.
+
+If you do not have a workstream ID yet, use local mode instead:
+
+```bash
+./.sdp/bin/sdp build "what you want to change" --dry-run --format text
+```
+
+That gives you a local draft path without requiring queue-backed execution.
+
 ## Fallback Mode
 
 If your Cursor runtime cannot spawn subagents, use the manual checklists in

.cursor/rules/beads.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/beads.md -->
-
 # /beads — Beads
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `beads` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/beads/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/bugfix.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/bugfix.md -->
-
 # /bugfix — Quality Bug Fixes
 
 When calling `/bugfix issue NNN`:
@@ -31,11 +30,11 @@ git status  # MUST show "up to date with origin"
 
 ## Quick Reference
 
-**Input:** P1/P2 issue  
+**Input:** P1/P2 issue
 **Output:** Bug fixed + tests + pushed to origin
 
 | Aspect | Hotfix | Bugfix |
 |--------|--------|--------|
 | Severity | P0 | P1/P2 |
 | Branch from | main | main |
-| Testing | Fast | Full |
+| Testing | Fast | Full |

.cursor/rules/build.mdc

@@ -5,20 +5,19 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/build.md -->
-
 # /build — Execute Workstream
 
 When calling `/build {WS-ID}`:
 
-1. Load skill: `@.claude/skills/build/SKILL.md`
-2. Run pre-build hook: `hooks/pre-build.sh {WS-ID}`
+1. Load skill: `build`
+2. Run pre-build hook: `scripts/hooks/pre-build.sh {WS-ID}`
 3. Read WS plan
 4. Execute steps using TDD
-5. Run post-build hook: `hooks/post-build.sh {WS-ID}`
+5. Run post-build hook: `scripts/hooks/post-build.sh {WS-ID}`
 6. Append Execution Report to WS file
 
 ## Quick Reference
 
 **Input:** `workstreams/backlog/WS-XXX-*.md`
 **Output:** Code + tests + Execution Report
-**Next:** `/build WS-XXX-02` or `/review F{XX}`
+**Next:** `/build WS-XXX-02` or `/review F{XX}`

.cursor/rules/ci-triage.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/ci-triage.md -->
-
 # /ci-triage — Ci-triage
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `ci-triage` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/ci-triage/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/codereview.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/codereview.md -->
-
 # /codereview — Review
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `review` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/review/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/debug.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/debug.md -->
-
 # /debug — Debug
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `debug` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/debug/SKILL.md`
-- Agents: `prompts/agents/planner.md`
+- Agents: `prompts/agents/planner.md`

.cursor/rules/deliver.mdc

@@ -33,4 +33,4 @@ Do **not** stop for routine fix/rebuild decisions. **Do** stop to escalate:
 - Ambiguous findings with no clear fix strategy
 - Phase-1 cap hit at cycle 5 (operator must paste deferred-P3 list into spin-out bead)
 
-See `.agents/skills/delivery-loop.md` for the full state machine and `docs/plans/2026-04-22-deliver-skill-review-design.md` for the design rationale.
+See `.agents/skills/delivery-loop.md` for the full state machine and `docs/plans/2026-04-22-deliver-skill-review-design.md` for the design rationale.

.cursor/rules/deploy.mdc

@@ -5,12 +5,11 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/deploy.md -->
-
 # /deploy — Deploy Feature
 
 When calling `/deploy {feature} [version_bump]`:
 
-1. Load skill: `.claude/skills/deploy/SKILL.md`
+1. Load skill: `deploy`
 2. Pre-flight: run quality gates (see AGENTS.md), verify APPROVED
 3. Version: bump semver (patch/minor/major)
 4. Generate: CHANGELOG, release notes
@@ -31,4 +30,4 @@ When calling `/deploy {feature} [version_bump]`:
 
 - `@deploy <feature-id>` — patch (0.5.0 → 0.5.1)
 - `@deploy <feature-id> minor` — minor (0.5.0 → 0.6.0)
-- `@deploy <feature-id> major` — major (0.5.0 → 1.0.0)
+- `@deploy <feature-id> major` — major (0.5.0 → 1.0.0)

.cursor/rules/design.mdc

@@ -5,12 +5,11 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/design.md -->
-
 # /design — Analyze + Plan
 
 When calling `/design {slug}`:
 
-1. Load full prompt: `@.claude/skills/design/SKILL.md`
+1. Load full prompt: `design`
 2. Read PROJECT_MAP.md and INDEX.md
 3. Read draft: `docs/drafts/idea-{slug}.md`
 4. Create all WS files in `workstreams/backlog/`
@@ -21,4 +20,4 @@ When calling `/design {slug}`:
 
 **Input:** `docs/drafts/idea-{slug}.md`
 **Output:** `docs/workstreams/backlog/WS-XXX-*.md`
-**Next:** `/build WS-XXX-01`
+**Next:** `/build WS-XXX-01`

.cursor/rules/feature.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/feature.md -->
-
 # /feature — Feature
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `feature` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/feature/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/hotfix.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/hotfix.md -->
-
 # /hotfix — Emergency Production Fixes
 
 When calling `/hotfix "description" --issue-id=001`:
@@ -32,12 +31,12 @@ git push origin main --tags
 
 ## Quick Reference
 
-**Input:** P0 CRITICAL issue  
+**Input:** P0 CRITICAL issue
 **Output:** Production fix + pushed to origin
 
 **Key Rules:**
 - Minimal changes only
 - No refactoring
 - No new features
 - Fast testing
-- Backport mandatory
+- Backport mandatory

.cursor/rules/idea.mdc

@@ -5,12 +5,11 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/idea.md -->
-
 # /idea — Requirements Gathering
 
 When calling `/idea {description}`:
 
-1. Load full prompt: `@.claude/skills/idea/SKILL.md`
+1. Load full prompt: `idea`
 2. Execute Mandatory Initial Dialogue
 3. Create draft in `docs/drafts/idea-{slug}.md`
 4. Output summary for user
@@ -19,4 +18,4 @@ When calling `/idea {description}`:
 
 **Input:** Feature description from user
 **Output:** `docs/drafts/idea-{slug}.md`
-**Next:** `/design idea-{slug}`
+**Next:** `/design idea-{slug}`

.cursor/rules/issue.mdc

@@ -5,12 +5,11 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/issue.md -->
-
 # /issue — Analyze & Route Issues
 
 When calling `/issue "description"`:
 
-1. Load full prompt: `@.claude/skills/issue/SKILL.md`
+1. Load full prompt: `issue`
 2. Systematic debugging (5 phases):
    - Symptom analysis
    - Hypothesis formation
@@ -33,4 +32,4 @@ When calling `/issue "description"`:
 - P2 MEDIUM → New WS
 - P3 LOW → Defer
 
-**Next:** `/hotfix` or `/bugfix` depending on severity
+**Next:** `/hotfix` or `/bugfix` depending on severity

.cursor/rules/oneshot.mdc

@@ -5,12 +5,11 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/oneshot.md -->
-
 # /oneshot — Autonomous Feature Execution
 
 When calling `/oneshot F{XX}` in Cursor:
 
-1. Load skill: `@.claude/skills/oneshot/SKILL.md`
+1. Load skill: `oneshot`
 2. Run `sdp-orchestrate --feature F{XX} --next-action` as the outer loop
 3. Execute each phase inline:
    - **build**: @build {ws_id} → commit → `sdp-orchestrate --feature F{XX} --advance --result <commit>`
@@ -21,4 +20,4 @@ When calling `/oneshot F{XX}` in Cursor:
 **Input:** Feature ID (from @feature or ROADMAP)
 **Output:** All WS executed + CI green. No "Next steps" or handoff lists.
 
-**opencode:** Use `sdp-orchestrate --feature F{XX} --runtime opencode` as the outer loop. opencode lacks Stop hooks — the outer loop CLI replaces them.
+**opencode:** Use `sdp-orchestrate --feature F{XX} --runtime opencode` as the outer loop. opencode lacks Stop hooks — the outer loop CLI replaces them.

.cursor/rules/prd.mdc

@@ -5,9 +5,8 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/prd.md -->
-
 # /prd
 
 PRD generation is now part of @vision. Use `@vision "project-name"` for initial PRD creation. Use `@vision "project-name" --update` to regenerate diagrams from @prd annotations.
 
-See `prompts/skills/vision/SKILL.md` for documentation.
+See `prompts/skills/vision/SKILL.md` for documentation.

.cursor/rules/protocol-consistency.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/protocol-consistency.md -->
-
 # /protocol-consistency — Protocol-consistency
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `protocol-consistency` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/protocol-consistency/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/prototype.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/prototype.md -->
-
 # /prototype — Prototype
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `prototype` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/prototype/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/reality-check.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/reality-check.md -->
-
 # /reality-check — Reality-check
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `reality-check` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/reality-check/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`

.cursor/rules/reality.mdc

@@ -5,7 +5,6 @@ alwaysApply: false
 ---
 <!-- GENERATED by sdp generate-adapters from sdp.manifest.yaml — do not edit manually -->
 <!-- Source: prompts/commands/reality.md -->
-
 # /reality — Reality
 
 ## Overview
@@ -32,4 +31,4 @@ The command delegates to the `reality` skill, which provides:
 ## Related
 
 - Skills: `prompts/skills/reality/SKILL.md`
-- Agents: `prompts/agents/builder.md`
+- Agents: `prompts/agents/builder.md`