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.
In this repository15 other files it gives its agents.
- pm-auditSkill
- pm-brainstormSkill
- pm-cleanupSkill
- pm-codesmellsSkill
- pm-doneSkill
- pm-exploreSkill
- pm-housekeepingSkill
- pm-implementSkill
- pm-nextSkill
- pm-planSkill
- pm-scaffoldSkill
- pm-statusSkill
- pm-triageSkill
- pm-troubleshootSkill
- pm-verifySkill
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
- Very specific: line 17
Call mcp__pinkrooster__get_project_status with projectPath set to the directory above
- Distinctive: line 19
If no project found, tell the user and offer to register it
- Very specific: line 23
Call mcp__pinkrooster__get_feature_request_details with the FR ID from $ARGUMENTS.
- Very specific: line 25
If no ID is provided, call mcp__pinkrooster__get_feature_requests with projectId and stateFilter: "Inactive" to list candidates, then use …
- Very specific: line 41
Use Serena's get_symbols_overview or list_dir to understand the relevant areas of the codebase mentioned in the FR description
- Distinctive: line 42
Check existing PM skills, patterns, and conventions if the FR relates to a new skill or workflow
- Very specific: line 49
Call mcp__pinkrooster__create_or_update_feature_request with projectId, featureRequestId, and improved fields:
- Distinctive: line 63
Keep the original author's intent — refine and expand, don't reimagine.
- Distinctive: line 66
If missing or weak, derive from the description:
- Distinctive: line 72
If missing, write 5-8 concrete, testable acceptance criteria numbered as a list.
- Distinctive: line 75
If missing or generic, classify:
- Distinctive: line 81
If missing, infer from business value and scope.
- Distinctive: line 90
Use concrete, specific language — not vague platitudes
- Distinctive: line 101
If they are well-written, keep them and only add missing perspectives
- Very specific: line 102
If they are vague or low-quality, use action: "Update" with the index to improve them
- Very specific: line 103
If they are redundant, use action: "Remove" with the index
- Very specific: line 137
Never change the FR's status — refinement is about content, not workflow state
- Distinctive: line 138
Never remove existing user stories unless they are clearly redundant or wrong — prefer updating
- Distinctive: line 139
Always read the FR details first before making any changes
- Distinctive: line 140
If the FR is already well-specified (detailed description, business value, acceptance criteria, user stories), tell the user and suggest …
- Distinctive: line 142
Always show the before/after summary so the user can verify the refinement
≈ “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.
Files
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_statuswithprojectPathset 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:
- Use Serena's
get_symbols_overvieworlist_dirto understand the relevant areas of the codebase mentioned in the FR description - Check existing PM skills, patterns, and conventions if the FR relates to a new skill or workflow
- Review similar existing FRs via
mcp__pinkrooster__get_feature_requeststo 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 capabilityEnhancement— extends existing functionalityImprovement— 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
goalshould describe observable behavior, not internal implementation - The
benefitshould 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 theindexto improve them - If they are redundant, use
action: "Remove"with theindex
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
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
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.

