메뉴
Use when the user asks to "review backlog", "analyze backlog quality", "backlog health", or invokes /compass-rose:backlog. Reads all open and wontfix issue files and analyzes quality via the backlog-analyzer agent.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | backlog |
| description | Use when the user asks to "review backlog", "analyze backlog quality", "backlog health", or invokes /compass-rose:backlog. Reads all open and wontfix issue files and analyzes quality via the backlog-analyzer agent. |
| allowed-tools | Bash, Read, Glob, Agent |
You are now in Backlog Review Mode. Your role is to read all active issue files, assess their quality and readiness, and recommend the top 2-3 items to work on next based on priority, size, and definition quality.
.lore/work/issues/*.html filesopen or wontfix (exclude done and closed)Use Glob to find all issue files:
.lore/work/issues/*.html
If no files exist, stop and say:
No issue files found in .lore/work/issues/.
To get started, create your first issue with /compass-rose:add-item.
For each file, use Read to extract the meta tags from the <head> and the body text:
Meta tags to extract:
priority — e.g., P0, P1, P2, P3size — e.g., S, M, L, XLstatus — e.g., open, wontfix, done, closeddate — creation datetitle — issue title (also available as <title> tag)Body text: Extract the readable text content from the HTML body.
Include items where status is open or wontfix. Exclude items where status is done or closed.
wontfix items are intentionally included. They are deferred, not done, and should surface if context changes.
If no items remain after filtering:
No open or deferred issues found in .lore/work/issues/.
All issues are either done or closed.
To add new work, use /compass-rose:add-item.
Build a JSON array from the filtered items:
[
{
"filepath": ".lore/work/issues/fix-login-timeout.html",
"title": "Fix login timeout",
"priority": "P1",
"size": "S",
"status": "open",
"date": "2026-05-01",
"body": "<p>Users experiencing timeouts when submitting the login form...</p>"
}
]
Invoke the backlog-analyzer agent with the prepared item data:
Analyze these project items and recommend the top 2-3 to work on next:
[Paste the formatted JSON array]
Focus on:
- Definition quality (clarity, completeness, acceptance criteria)
- Priority distribution
- Size balance (prefer a mix of quick wins and substantial work)
- Overall backlog health
Return your analysis with detailed rationale for each recommendation.
Agent Reference: compass-rose/agents/backlog-analyzer.md
The agent will:
Display the agent's analysis output directly to the user. The agent returns structured markdown with:
# Backlog Analysis Results
**Items Analyzed**: [N total]
**Well-Defined Items**: [X items with score 8-10]
**Items Needing Clarification**: [Y items with score <5]
## Top Recommendations
### Recommendation 1: [Title] ([filename])
**Priority**: [P0/P1/P2/P3] | **Size**: [S/M/L/XL] | **Definition Quality**: [Well-Defined/Defined/Vague/Poorly Defined] ([score]/10)
**Rationale**:
- [Why this is recommended - link priority, size, definition quality]
- [What makes it ready to work on]
- [Any specific strengths]
**Definition Assessment**:
- **Clarity** ([0-3]/3): [Brief assessment]
- **Completeness** ([0-3]/3): [Brief assessment]
- **Acceptance Criteria** ([0-4]/4): [Brief assessment]
**File**: [filepath]
---
### Recommendation 2: [Title] ([filename])
[Same structure as Recommendation 1]
---
### Recommendation 3: [Title] ([filename]) [OPTIONAL]
[Same structure as Recommendation 1]
---
## Backlog Health Summary
**Priority Distribution**: [X P0, Y P1, Z P2, W P3]
**Size Distribution**: [A S, B M, C L, D XL]
**Definition Quality**:
- Well-Defined (8-10): [X items]
- Defined (5-7): [Y items]
- Vague (2-4): [Z items]
- Poorly Defined (0-1): [W items]
**Observations**:
- [Notable patterns]
- [Quality trends]
- [Recommendations for backlog improvement]
## Items Needing Clarification
1. **[Title]** ([filename]) - Score: [X]/10
- Missing: [What needs to be added]
- Suggest: [How to improve definition]
All Items Poorly Defined:
If the backlog-analyzer agent reports all items have low quality scores (<5), still present the top 2-3 but emphasize the need for clarification:
Warning: Most backlog items lack sufficient detail.
The recommendations below are based on priority and size only. Each item needs
clarification before implementation can begin. Consider adding:
- Clear problem or feature descriptions
- Reproduction steps (for bugs) or use cases (for features)
- Explicit acceptance criteria
Recommended next step: Clarify the highest-priority items before starting work.
Missing Priority or Size Fields:
Warning: Some items are missing priority or size fields.
Recommendations for those items will fall back to definition quality and date.
Consider setting priority and size in each issue file's meta tags.
This skill implements the following specification requirements:
.lore/work/issues//compass-rose:next-item - Get an immediate recommendation for next work item (faster, simpler)/compass-rose:add-item - Create a new issue file/compass-rose:reprioritize - Codebase-aware priority updates| Aspect | /compass-rose:next-item | /compass-rose:backlog |
|---|---|---|
| Speed | Fast | Slower |
| Scope | Ready items only | All open and wontfix items |
| Analysis | Simple priority sort | Deep quality assessment |
| Output | Quick recommendation | Comprehensive backlog health report |
| Use Case | "What's next?" | "What's the state of my backlog?" |
compass-rose/agents/backlog-analyzer.md (quality scoring and recommendation)Use when checking whether field-guide reference pages still match current code and tests. Reads pages plus fg-evidence anchors, reports or fixes semantic drift, and can use sub-agents for parallel audit batches. Triggers include "resolve drift", "semantic drift", "check accuracy", "verify the reference against code", and "audit the wiki".
Use after ingesting or editing field-guide pages to attach living code/test anchors. Adds or refreshes fg-evidence metadata without trying to prove prose accuracy. Triggers include "update evidence", "wire evidence", "add evidence anchors", and "connect the wiki to code".
Use when extracting knowledge from .lore/ artifacts into the reference wiki. Converts work artifacts (retros, specs, plans, research, notes) into structured wiki pages grouped by knowledge type. Reads Markdown or HTML sources and writes Markdown by default. Triggers include "ingest this", "extract knowledge from", "add this to the wiki", "update the reference from", and "populate the field guide".
Use when setting up field-guide on a project for the first time, or to refresh the scheduled lint job after it expires. Triggers include "init field-guide", "set up field-guide", "refresh field-guide lint schedule", and "bootstrap the wiki".
This skill is used when the project wiki needs a health check. Triggers include 'lint the wiki', 'check wiki health', 'run field-guide lint', 'are there stale wiki pages', 'find orphaned pages', 'check for contradictions in the wiki'.
Use when asking questions about the project wiki, querying accumulated knowledge, searching project decisions or lessons. Triggers include "query the wiki", "what does the wiki say about", "search project knowledge", and natural language questions directed at the field-guide wiki.