| name | plan-2b-v2-prep-issue |
| description | Generate terse, industry-standard issue text from specs and plans for Azure DevOps, GitHub Issues, or any tracker. Domain-aware v2. |
Please deep think / ultrathink as this is a complex task.
plan-2b-v2-prep-issue
Generate concise, well-structured issue text from the feature specification and plan artifacts. This command creates terse, actionable issue content suitable for any issue tracker (GitHub, Azure DevOps, Jira, etc.).
Purpose: Extract clarity from complex specs. The issue is a signpost, not a replacement for the spec.
🚫 NO TIME ESTIMATES
Use Complexity Score (CS 1-5) only — no hours, days, or story points.
User input:
$ARGUMENTS
# Optional flags:
# --phase N # Generate Story/Task level issue for phase N (requires /the-flow 5 output)
# --type TYPE # Override auto-detected type: feature|story|task
## Workflow
1) **Resolve plan folder and artifacts**:
- Parse user input for plan folder path or feature slug
- PLAN_DIR = `docs/plans/<ordinal>-<slug>/`
- Locate available artifacts:
* SPEC_FILE = `${PLAN_DIR}/<slug>-spec.md` (REQUIRED)
* PLAN_FILE = `${PLAN_DIR}/<slug>-plan.md` (optional)
* TASKS_DIR = `${PLAN_DIR}/tasks/` (optional)
- If SPEC_FILE not found: ERROR "Spec not found. Run /the-flow 1b first."
2) **Load domain context** (additive — skip if no domains):
- If `docs/domains/registry.md` exists → read registered domains
- If `docs/domains/domain-map.md` exists → read domain topology
- Read `## Target Domains` from SPEC_FILE (if present) → DOMAIN_LIST
- If PLAN_FILE exists → read `## Target Domains` and `## Domain Manifest` tables
- For each domain in DOMAIN_LIST → read `docs/domains/<slug>/domain.md` for:
* Concepts, contracts, dependencies
- Identify CROSS_DOMAIN_DEPS: domains that depend on each other per domain map
- HAS_DOMAINS = true if any domain info found, false otherwise
3) **Determine issue type** (auto-detect or use --type flag):
- If --phase N provided AND `${TASKS_DIR}/phase-N-*/tasks.md` exists:
* TYPE = story (or task if --type task)
* PHASE_DOSSIER = the tasks.md file
- Else if PLAN_FILE exists:
* TYPE = feature (plan context available)
- Else:
* TYPE = feature (spec-only)
4) **Extract content from artifacts**:
**From SPEC_FILE** (always read):
- TITLE = H1 heading text
- SUMMARY = ## Summary section (2-3 sentences)
- GOALS = ## Goals section (bullet points)
- NON_GOALS = ## Non-Goals section (bullet points)
- AC = ## Acceptance Criteria section (numbered list)
- COMPLEXITY = ## Complexity section (CS score and breakdown)
- RISKS = ## Risks & Assumptions section (key risks only)
- TARGET_DOMAINS = ## Target Domains section (if present)
**From PLAN_FILE** (if exists and TYPE = feature):
- CRITICAL_FINDINGS = top 3 from Key Findings section
- PHASE_COUNT = number of Implementation Phases
- ADR_REFS = any ADR references
- DOMAIN_MANIFEST = ## Domain Manifest table (file → domain mapping)
**From PHASE_DOSSIER** (if TYPE = story/task):
- PHASE_TITLE = Phase heading
- PHASE_OBJECTIVE = from context brief / Purpose
- PHASE_TASKS = task count from tasks table
- PHASE_AC = derive from task Success Criteria column (observable outcomes)
- PHASE_DOMAIN = Domain column from task table (primary domain for this phase)
5) **Generate issue content**:
**For TYPE = feature**:
```markdown
# [TITLE]
**Type**: Feature
**Complexity**: [COMPLEXITY score] ([S,I,D,N,F,T breakdown])
[If HAS_DOMAINS:]
**Domain**: [primary domain from TARGET_DOMAINS, or "multiple — see Domain Impact"]
## Objective
[SUMMARY - 2-3 sentences on WHAT and WHY]
## Acceptance Criteria
[AC - numbered, testable criteria from spec]
## Goals
[GOALS - bullet points]
## Non-Goals
[NON_GOALS - bullet points, keep terse]
## Context
[If PLAN_FILE exists: "[PHASE_COUNT] implementation phases planned. See plan for details."]
[If CRITICAL_FINDINGS: Brief mention of top constraint/finding]
[If ADR_REFS: "Key decisions documented in: [ADR links]"]
[If HAS_DOMAINS AND CROSS_DOMAIN_DEPS:]
### Cross-Domain Dependencies
[List domain pairs with dependency direction and contract type, one line each.
Example: "`billing` → `auth` (consumes: user identity contract)"]
[If HAS_DOMAINS:]
## Domain Impact
| Domain | Status | Relationship | Changes |
|--------|--------|-------------|---------|
[From TARGET_DOMAINS table — one row per domain touched]
[If DOMAIN_MANIFEST exists: "See plan Domain Manifest for full file mapping."]
## Key Risks
[Top 2-3 risks from RISKS section, one line each]
## Labels
[If HAS_DOMAINS: Suggest domain labels, e.g., `domain:auth`, `domain:billing`]
[Always suggest type label: `type:feature`]
[If COMPLEXITY: suggest complexity label, e.g., `complexity:cs-3`]
## References
- Spec: `[relative path to SPEC_FILE]`
[If PLAN_FILE exists:]
- Plan: `[relative path to PLAN_FILE]`
---
*Generated from spec. See referenced documents for implementation details.*
For TYPE = story:
# [PHASE_TITLE]
**Type**: Story
**Parent**: [TITLE] (Feature)
**Phase**: [N] of [PHASE_COUNT]
[If HAS_DOMAINS:]
**Domain**: [PHASE_DOMAIN — primary domain for this phase]
## Objective
[PHASE_OBJECTIVE from context brief]
## Acceptance Criteria
[PHASE_AC - derived from task success criteria, numbered]
## Scope
- Tasks: [PHASE_TASKS count]
- [Brief scope from phase deliverables]
## Non-Goals (This Phase)
[From phase brief Non-Goals section if exists]
[If HAS_DOMAINS AND phase touches multiple domains:]
## Cross-Domain Notes
[Which other domains this phase touches and why — one line each]
## Labels
[If HAS_DOMAINS: `domain:[PHASE_DOMAIN]`]
`type:story`, `phase:[N]`
## References
- Spec: `[relative path to SPEC_FILE]`
- Plan: `[relative path to PLAN_FILE]`
- Phase Dossier: `[relative path to PHASE_DOSSIER]`
---
*Generated from phase dossier. See referenced documents for task details.*
For TYPE = task:
# [Task description from tasks table]
**Type**: Task
**Parent**: [PHASE_TITLE] (Story)
**Task ID**: [T00N]
[If HAS_DOMAINS:]
**Domain**: [domain from task table Domain column]
## Objective
[Task description with context]
## Done When
[Success criteria from tasks table]
## Dependencies
[From Dependencies column, or "None"]
[If HAS_DOMAINS and task depends on another domain's contract:]
⚠️ Cross-domain: depends on `[other-domain]` contract — coordinate changes.
## Labels
[If HAS_DOMAINS: `domain:[task-domain]`]
`type:task`
## References
- Phase Dossier: `[relative path to PHASE_DOSSIER]`
---
*Generated from task dossier.*
-
Save issue file:
- Create
${PLAN_DIR}/issues/ directory if not exists
- ISSUE_SLUG = generate from title
- For feature:
${PLAN_DIR}/issues/feature-[ISSUE_SLUG].md
- For story:
${PLAN_DIR}/issues/story-phase-[N]-[ISSUE_SLUG].md
- For task:
${PLAN_DIR}/issues/task-[TASK_ID]-[ISSUE_SLUG].md
- Write generated content to file
-
Present output:
- Display the generated issue content
- Show the saved file path
- Remind user: "Copy to your issue tracker or request additional issues."
Gates
- SPEC_FILE must exist
- Generated content must be terse (signpost, not duplication)
- All paths in References must be relative (repo-portable)
- No platform-specific fields (neutral markdown)
- No time estimates — CS scores only
Success Message
✅ Issue generated: [relative path to issue file]
Type: [feature|story|task]
Title: [TITLE]
Complexity: [CS score]
[If HAS_DOMAINS: "Domains: [list]"]
[If HAS_DOMAINS: "Suggested labels: [domain labels]"]
[Display generated issue content]
---
Copy the above to your issue tracker, or:
- Run again with --phase N for story-level issues
- Request additional issues for other phases
Integration Notes
- Can run after /the-flow 1b (spec only → feature issue)
- Can run after /the-flow 3 (with plan context → richer feature issue with domain impact)
- Can run after /the-flow 5 (with --phase N → domain-aware story issues)
- Multiple runs accumulate in issues/ folder (one file per issue)
- Domain awareness is additive — issues generate correctly without domains
