feature

name	feature
description	Feature planning orchestrator (discovery -> idea -> ux -> design -> workstream tree)
version	9.0.0
depends_on	@discovery v1
changes	[{"v9":"Added --design-only (alias for --quick, explicit naming)"},{"v8":"Full product discovery flow with @discovery, @ux, impact analysis"},"Added --quick (skip @discovery), --infra (skip @ux)",{"Step 3.5":"Impact analysis after @design"}]
examples	["@feature 'Add user authentication' --default # Full interactive pipeline","@feature 'Add user authentication' --quick # Skip to @design only, 0 questions","@feature 'Add user authentication' --design-only # Same as --quick, explicit name","@feature 'Add payment processing' --auto # Non-interactive, from roadmap/plan docs"]

@feature

Orchestrate product discovery, requirements, UX research, and workstream design.

Phase 0: This skill targets Go projects (e.g. go build/go test in acceptance criteria). Language-agnostic expansion is planned.

Three Explicit Modes (No Auto-Detection)

Key Principle: User must explicitly choose the mode. No hidden context detection. Same input + same mode = identical output.

Mode	When to use	Steps	Questions
--default (or no flag)	New/exploratory feature. Full interactive discovery.	0, 1, 2, 2.5, 3, 3.5, 4	Interactive (3-5 questions)
--quick / --design-only	User knows what they want, just needs workstreams.	3 only	0 questions - goes directly to @design
--auto	Feature already described in roadmap/plan. Non-interactive.	0, 3, 4 only	0 questions - reads from docs/ROADMAP.md

Mode Behavior Guarantee

Deterministic: Each mode produces identical behavior given identical input. No context sniffing, no heuristics, no "smart defaults."

• --default: Always asks the same questions in the same order
• --quick / --design-only: Always skips to @design with zero questions
• --auto: Always reads from roadmap/plan docs, never asks questions

---

--auto Mode (Recommended for Roadmap Features)

For features already defined in docs/ROADMAP.md:

Step A: Extract from Roadmap (Non-Interactive)

1. Find the feature in the roadmap: rg "F0\d\d" docs/ROADMAP.md -A 10
2. Extract: feature ID, description, success criteria, listed deliverables
3. Identify scope: what files/packages this touches (from deliverables and codebase)

No questions asked. Read from docs only.

Step B: Auto-Generate Workstreams

For each deliverable, create workstream files using this format:

---
ws_id: 00-FFF-SS
feature_id: FFFF
status: open
priority: P1
size: M
depends_on: []
ws_kind: leaf|aggregate
parent_ws_id: null|00-FFF-SS
dispatch_lifecycle: active
---

# 00-FFF-SS: Feature Name — Step Description

## Goal
One paragraph: what this workstream does and why.

## Scope Files
- path/to/file/or/dir (exact files or directory prefixes)
- ...

## Beads
- primary: sdplab-XXXX      # leaf only
- finding: sdplab-YYYY      # optional

## Acceptance Criteria
- [ ] Specific, testable criterion 1
- [ ] Specific, testable criterion 2
- [ ] go build ./... passes
- [ ] go test ./... passes

Rules:

• aggregate = container, no primary issue
• leaf = executable, may have one primary issue
• parent_ws_id links leaf to aggregate (max 1 nesting layer)

Step C: Create Beads Issues

For each leaf workstream:

bd create --title="WS FFF-SS: Short title" --type=task

Update .beads-sdp-mapping.jsonl:

{"sdp_id":"00-FFF-SS","beads_id":"sdp_dev-XXXX","updated_at":"2026-..."}

Step D: Report

Output: feature ID, workstream count, file names, Beads IDs, ready-to-run command (@build 00-FFF-01 or @oneshot F0FF).

---

--quick / --design-only Mode (@design Only, Zero Questions)

For users who know what they want and just need workstreams. Zero questions, deterministic.

Steps: Run @design directly, produce workstream files, skip roadmap/UX/impact analysis.

When to use: Clear feature description, no product research needed, immediate workstreams required.

---

--default Mode (Full Interactive Pipeline)

Standard mode for new/exploratory features. Full discovery with interactive questions.

Step 0: Roadmap Pre-Check

Use @discovery "feature description" or manually: extract keywords, rg docs for overlap, present Overlap Report (HIGH/MEDIUM), gate on user resolution.

Step 1: Quick Interview (3-5 questions)

Problem, Users, Success. Gate: if vague (<200 words), ask clarification. Use @discovery output if available.

Step 2: @idea

@idea "..." --spec docs/drafts/discovery-{slug}.md (use @discovery output if Step 0 ran)

Step 2.5: @ux — unless --infra

Auto-trigger when @idea output has user-facing keywords (ui, user, interface, dashboard, form) and lacks infra (K8s, CRD, CLI-only).

Step 3: @design

@design {task_id} — workstream files in docs/workstreams/backlog/ using Workstream file format above.

Step 3.5: Impact Analysis

Read scope files, grep/rg for conflicts. Categorize: FILE CONFLICT, DATA BOUNDARY, DEPENDENCY CHAIN, PRIORITY SHIFT. Present report, user acknowledges.

Step 4: Verify Outputs

Check discovery brief, idea spec, ux output, workstreams exist, direct execution targets are leaf workstreams.

---

Key Principle: Protocol is Invisible

User sees: feature description → workstreams created → ready to build. Workstream files, scope declarations, beads IDs are plumbing.

Write Plan (F101)

Before creating workstream files and docs, emit a write plan:

1. Enumerate — List every file the skill will CREATE / MODIFY / DELETE with a one-line reason. Covers workstream files, discovery briefs, idea specs, UX outputs, and beads mappings.
2. Flags:
   • --dry-run — Emit write plan only. Do NOT create, modify, or delete any file.
   • --yes — Skip confirmation prompt. Execute immediately. Intended for CI/non-interactive.
3. Confirm — Present the plan to the user and wait for explicit approval (unless --yes).
4. Log — Append write plan event to .sdp/log/events.jsonl (sanitize file paths before logging: strip newlines, ensure valid JSON escaping):

   {"spec_version":"v1.0","event_id":"<uuid>","timestamp":"<ISO-8601>","source":{"system":"sdp-lab","component":"feature"},"event_type":"decision.made","payload":{"decision_type":"write_plan","plan":[{"path":"...","action":"CREATE|MODIFY|DELETE","reason":"..."}]},"context":{"feature_id":"<F-id if known>","workstream_id":"<ws-id if applicable>"}}
   

   Include context fields only when the ID is known at plan time. Omit unavailable fields rather than inventing placeholders.

   Note: Phase 1 uses prompt-level write boundaries (CLI out of scope). Aligns with schema/contracts/orchestration-event.schema.json via event_type: "decision.made". Phase 2 CLI will emit natively.

Output format:

WRITE PLAN for @feature <feature-id>:
  CREATE: path/to/new/file — <reason>
  MODIFY: path/to/existing/file — <reason>
  DELETE: path/to/removed/file — <reason>

Proceed? [y/n]

Write-plan flags:

• No flag: Show plan → Confirm → Execute
• --dry-run: Show plan → STOP
• --yes: Show plan → Execute immediately (no prompt)

Note: --dry-run and --yes are orthogonal to skill mode flags (--default, --quick, --auto). They can be combined with any mode (e.g. @feature "X" --quick --dry-run).

Completion

When all workstreams are created and verified, output:

@feature complete. Feature {ID}: {count} workstreams created.
  Aggregate: 00-{FFF}-00
  Leaves: 00-{FFF}-01 .. 00-{FFF}-{NN}

Next: @build 00-{FFF}-01  or  @oneshot F{XX}

Recovery

Symptom	Fix
Skill produces no output	Run @init first to scaffold project structure; then re-run @feature
"checkpoint not found"	Run sdp-orchestrate --feature <ID> to create initial checkpoint
"workstream files missing"	Run sdp-orchestrate --index to verify, then @feature to regenerate
Skill hangs / no progress	Check .sdp/log/events.jsonl for last event; use sdp reset --feature <ID> if stuck
Review loop exceeds 3 rounds	Use @review --override "reason", @review --partial, or @review --escalate
--design-only produces no workstreams	Run @init to scaffold project structure, then re-run with --design-only

See Also

@discovery — Product discovery gate | @idea — Requirements | @ux — UX research | @design — Workstream planning | @build — Execute leaf workstream | @oneshot — Execute all ready leaf workstreams