agentleFS
pinkroosteraipinkroosterai/PinkRoosterMcp/.claude/skills/pm-refine-fr/SKILL.md

pm-refine-fr

Skill2 starsChanged 8 months agoMITView on GitHub

Refine a feature request by analyzing its current state, rewriting the description with full detail, filling in missing fields (business value, acceptance summary, category, priority), and adding structured user stories.

The file

Show its rules and text

We split this skill into the rules it gives an agent

21 rules. About 0 are common, 13 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈13

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Call mcp__pinkrooster__get_project_status with projectPath set to the directory above

    line 17
  2. Distinctive:

    If no project found, tell the user and offer to register it

    line 19
  3. Very specific:

    Call mcp__pinkrooster__get_feature_request_details with the FR ID from $ARGUMENTS.

    line 23
  4. Very specific:

    If no ID is provided, call mcp__pinkrooster__get_feature_requests with projectId and stateFilter: "Inactive" to list candidates, then use …

    line 25
  5. Very specific:

    Use Serena's get_symbols_overview or list_dir to understand the relevant areas of the codebase mentioned in the FR description

    line 41
  6. Distinctive:

    Check existing PM skills, patterns, and conventions if the FR relates to a new skill or workflow

    line 42
  7. Very specific:

    Call mcp__pinkrooster__create_or_update_feature_request with projectId, featureRequestId, and improved fields:

    line 49
  8. Distinctive:

    Keep the original author's intent — refine and expand, don't reimagine.

    line 63
  9. Distinctive:

    If missing or weak, derive from the description:

    line 66
  10. Distinctive:

    If missing, write 5-8 concrete, testable acceptance criteria numbered as a list.

    line 72
  11. Distinctive:

    If missing or generic, classify:

    line 75
  12. Distinctive:

    If missing, infer from business value and scope.

    line 81
  13. Distinctive:

    Use concrete, specific language — not vague platitudes

    line 90
  14. Distinctive:

    If they are well-written, keep them and only add missing perspectives

    line 101
  15. Very specific:

    If they are vague or low-quality, use action: "Update" with the index to improve them

    line 102
  16. Very specific:

    If they are redundant, use action: "Remove" with the index

    line 103
  17. Very specific:

    Never change the FR's status — refinement is about content, not workflow state

    line 137
  18. Distinctive:

    Never remove existing user stories unless they are clearly redundant or wrong — prefer updating

    line 138
  19. Distinctive:

    Always read the FR details first before making any changes

    line 139
  20. Distinctive:

    If the FR is already well-specified (detailed description, business value, acceptance criteria, user stories), tell the user and suggest …

    line 140
  21. Distinctive:

    Always show the before/after summary so the user can verify the refinement

    line 142

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Refine Feature Request

Analyze an existing feature request, rewrite it with comprehensive detail, fill in gaps, and add user stories.

Step 1: Resolve Project

  • Current directory: !pwd
  • Call mcp__pinkrooster__get_project_status with projectPath set to the directory above
  • Extract the projectId
  • If no project found, tell the user and offer to register it

Step 2: Load Feature Request

Call mcp__pinkrooster__get_feature_request_details with the FR ID from $ARGUMENTS.

If no ID is provided, call mcp__pinkrooster__get_feature_requests with projectId and stateFilter: "Inactive" to list candidates, then use the AskUserQuestion tool to let the user pick:

  • Question: "Which feature request would you like to refine?"
  • Header: "Select FR"
  • Options: Build from candidates (up to 4), e.g. [{label: "{frId}", description: "\"{name}\" ({status}, {priority})"}, ...]

Record the current state of all fields:

  • name, description, category, priority, status
  • businessValue, acceptanceSummary, requester
  • userStories (count and content)
  • attachments
  • linkedWorkPackages

Step 3: Analyze the Codebase for Context

To write a well-informed FR, understand the project context:

  1. Use Serena's get_symbols_overview or list_dir to understand the relevant areas of the codebase mentioned in the FR description
  2. Check existing PM skills, patterns, and conventions if the FR relates to a new skill or workflow
  3. Review similar existing FRs via mcp__pinkrooster__get_feature_requests to understand the level of detail expected and avoid overlap

This step is proportional to the FR's scope — a small enhancement needs less exploration than a new major feature.

Step 4: Rewrite the Feature Request

Call mcp__pinkrooster__create_or_update_feature_request with projectId, featureRequestId, and improved fields:

Name

  • Make it descriptive and specific (not generic like "Test FR")
  • Pattern: {feature-name}: {brief qualifier} (e.g., "pm-explore: AI-Driven Feature Discovery Skill")

Description

Expand the description into a comprehensive specification:

  • What: Clear explanation of what the feature does
  • Workflow: Numbered steps showing how the feature operates end-to-end
  • Configuration: Any settings, arguments, or options
  • Integration points: Which existing systems/tools/files it touches (reads and writes)
  • Explicit exclusions: What this feature is NOT (prevents scope creep)

Keep the original author's intent — refine and expand, don't reimagine.

Business Value

If missing or weak, derive from the description:

  • Who benefits and why?
  • What problem does this solve?
  • What becomes possible that wasn't before?

Acceptance Summary

If missing, write 5-8 concrete, testable acceptance criteria numbered as a list.

Category

If missing or generic, classify:

  • Feature — entirely new capability
  • Enhancement — extends existing functionality
  • Improvement — makes existing functionality better without adding new capabilities

Priority

If missing, infer from business value and scope. Don't change existing priority without flagging it.

Step 5: Add User Stories

Analyze the refined description to identify distinct user roles and capabilities, then add user stories via mcp__pinkrooster__manage_user_stories with action: "Add".

Guidelines for user stories:

  • Aim for 3-6 stories per FR (enough to cover the feature, not so many they become redundant)
  • Each story should represent a distinct user role OR a distinct capability
  • Use concrete, specific language — not vague platitudes
  • The goal should describe observable behavior, not internal implementation
  • The benefit should explain the "so what?" — why does the user care?

Common role patterns:

  • Primary user of the feature (the one who triggers it)
  • Secondary beneficiary (someone who benefits from the output)
  • Maintainer/admin (someone who configures or manages it)
  • Stakeholder (someone who reviews or approves results)

If user stories already exist, review them:

  • If they are well-written, keep them and only add missing perspectives
  • If they are vague or low-quality, use action: "Update" with the index to improve them
  • If they are redundant, use action: "Remove" with the index

Step 6: Present Summary

Show the user what changed:

## Refined: {frId} "{new name}"

### Fields Updated
| Field | Before | After |
|-------|--------|-------|
| Name | {old} | {new} |
| Description | {length}ch → {length}ch | Expanded |
| Business Value | {empty/existed} | {set/updated} |
| Acceptance Summary | {empty/existed} | {set/updated} |
| Category | {old} | {new or unchanged} |
| Priority | {old} | {new or unchanged} |

### User Stories ({count})
| # | Role | Goal | Benefit |
|---|------|------|---------|
| 1 | {role} | {goal} | {benefit} |
| ... | | | |

### Next Steps
- Scaffold a work package: `/pm-scaffold {frId}`
- View project status: `/pm-status`
- Triage and prioritize: `/pm-triage`

Constraints

  • Preserve the original author's intent — enhance, don't replace the core idea
  • Never change the FR's status — refinement is about content, not workflow state
  • Never remove existing user stories unless they are clearly redundant or wrong — prefer updating
  • Always read the FR details first before making any changes
  • If the FR is already well-specified (detailed description, business value, acceptance criteria, user stories), tell the user and suggest only minor improvements rather than rewriting
  • Cross-reference existing FRs to avoid suggesting overlapping scope in the description
  • Always show the before/after summary so the user can verify the refinement

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.