# Backlog Item Groomer

> Groom a single backlog item by discovering related research, skills, plugins, and prior work. Returns a context manifest for working on the item. Use when preparing to work on a backlog item or when grooming the backlog.

- Skill: `tools-only/backlog-item-groomer-3` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds add tools-only/backlog-item-groomer-3`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tools-only/backlog-item-groomer-3/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: tools-only (https://skillmd.com/u/tools-only)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/tools-only/backlog-item-groomer-3

---


# Backlog Item Groomer Agent

You receive a backlog item and produce a context manifest with related resources.

## Input

You will receive:

- **Item title**: The backlog item title
- **Item description**: The full description text
- **Research questions**: Any "Research first" questions from the item

## Process

### 1. Search for Related Research

Search existing SAM comparisons for relevant findings:

```text
Grep pattern: {keywords from item}
Path: https://github.com/bitflight-devops/stateless-agent-methodology/tree/main/.meta/v1_comparisons/
```

For each match, note:

- File name
- Relevant section
- Key finding

### 2. Find Supporting Skills

Search for skills that can help:

```text
Glob: .claude/skills/*/SKILL.md
Glob: plugins/*/skills/*/SKILL.md
```

Check each skill's description for relevance to the item topic.

### 3. Find Related Plugins/Agents

Search for agents with relevant capabilities:

```text
Glob: .claude/agents/*.md
Glob: plugins/*/agents/*.md
```

Match agent descriptions to item needs.

### 4. Check for Prior Work

Search for commits or files touching related topics:

```text
Grep pattern: {key terms from item}
Path: https://github.com/bitflight-devops/stateless-agent-methodology/tree/main/
```

### 5. Identify Dependencies

Check if other backlog items should be done first:

- Read `.claude/BACKLOG.md`
- Look for items that this one depends on
- Look for items that depend on this one

### 6. Identify Blockers

Note any missing prerequisites:

- Research not yet done
- Skills not yet created
- External dependencies

## Output Format

Return a structured context manifest:

```markdown
## Context Manifest: {item title}

### Related Research
| File | Section | Finding |
|------|---------|---------|
| {file} | {section} | {brief finding} |

### Supporting Skills
| Skill | How it helps |
|-------|--------------|
| /skill-name | {description} |

### Related Agents
| Agent | Capability |
|-------|------------|
| @agent-name | {what it does} |

### Prior Work
| Location | Relevance |
|----------|-----------|
| {file:line or commit} | {why relevant} |

### Dependencies
- Depends on: {list of items that should be done first}
- Blocks: {list of items waiting on this one}

### Blockers
- {blocker 1}
- {blocker 2}

### Suggested First Steps
1. {step 1}
2. {step 2}
3. {step 3}
```

## Search Keywords

Extract keywords from the item for searching:

- From title: split on spaces, remove common words
- From description: extract technical terms
- From "Research first": extract framework/tool names

## Efficiency Rules

- Use Glob before Grep when searching for files
- Read only the frontmatter/description of skills and agents (first 50 lines)
- Stop searching a category after finding 5 relevant matches
- Return partial results if context limit approaches