AgentSkillsCN

Feature Request Analyzer Skill

功能需求分析技能

SKILL.md

Feature Request Analyzer Skill

Purpose

Analyze the Feature Request dashboard in Notion to identify patterns, high-value opportunities, and prioritization insights.

Configuration

Read database settings from .claude/config/organization.json:

  • integrations.notion.feature_requests_database_id - Feature requests database
  • teams - Available teams for workload analysis
  • partners - Partners to track request volume
  • domains - Work domains for categorization

User Profile

If .claude/user-profile.json exists, read it to personalize the analysis:

  • domains: Highlight items in the user's domains first
  • team: Show team workload for the user's team prominently

Activation

  • User asks to "analyze feature requests"
  • User asks about "prioritization" or "what should we work on"
  • User asks to "review the backlog"

Data Source

Configure in organization.json under integrations.notion.feature_requests_database_id

Database Schema

PropertyTypeValues
TitletitleFree text
ImpactselectHigh, Medium, Low
StatusstatusNot started, In progress, Done
Domainmulti_select(from config domains)
Partner Requestedmulti_select(from config partners)
Teammulti_select(from config teams)
Roadmapped?checkboxYes/No
Use Case / Problem StatementtextFree text
Submission Datecreated_timeAuto
Submitted Bycreated_byAuto

Analysis Workflow

Step 1: Pull Current Data

Query open items (Status != 'Done'):

sql
SELECT Title, "Partner Requested", Domain, Impact, Status, Team, "Roadmapped?", "Use Case / Problem Statement"
FROM "collection://[feature_requests_database_id from config]"
WHERE Status != 'Done'

Step 2: Run Standard Analyses

Always produce these views:

A. Overview Stats

  • Total open requests
  • Roadmapped count and percentage
  • Unassigned to team count and percentage

B. High-Impact by Domain

Count requests grouped by Domain and Impact. Identify which domains have the most high-impact items.

sql
SELECT Domain, Impact, COUNT(*) as count
FROM "collection://[feature_requests_database_id]"
WHERE Status != 'Done'
GROUP BY Domain, Impact

C. Partner Request Volume

Identify which partners are requesting the most, and their impact distribution.

sql
SELECT "Partner Requested", COUNT(*) as request_count,
  SUM(CASE WHEN Impact = 'High' THEN 1 ELSE 0 END) as high_impact,
  SUM(CASE WHEN Impact = 'Medium' THEN 1 ELSE 0 END) as medium_impact
FROM "collection://[feature_requests_database_id]"
WHERE Status != 'Done' AND "Partner Requested" IS NOT NULL
GROUP BY "Partner Requested"
ORDER BY request_count DESC

D. Multi-Partner Requests (Highest Leverage)

Find items requested by multiple partners - these have highest ROI.

sql
SELECT Title, "Partner Requested", Impact, Domain, "Roadmapped?", Team
FROM "collection://[feature_requests_database_id]"
WHERE Status != 'Done' AND "Partner Requested" LIKE '%,%'
ORDER BY Impact DESC

E. Unroadmapped High-Impact (Quick Win Candidates)

High-impact items not yet on roadmap - potential quick wins or gaps.

sql
SELECT Title, "Partner Requested", Impact, Domain, Team
FROM "collection://[feature_requests_database_id]"
WHERE Status != 'Done' AND Impact = 'High' AND "Roadmapped?" = '__NO__'

F. Team Workload Distribution

Understand team capacity and backlog distribution.

sql
SELECT Team, Impact, COUNT(*) as count
FROM "collection://[feature_requests_database_id]"
WHERE Status != 'Done' AND Team IS NOT NULL
GROUP BY Team, Impact

Step 3: Synthesize Findings

After running queries, produce:

  1. Key Patterns: What themes emerge? Which domains are overloaded? Which partners are underserved?

  2. Top Recommendations:

    • Immediate wins (high impact, unroadmapped, multi-partner)
    • Internal efficiency gains (high impact, internal requests)
    • Roadmap validation (are roadmapped items actually prioritized?)
  3. Suggested Actions:

    • Team assignments needed
    • Items to consider for roadmap
    • Briefs to create

Output Format

Present findings in clear tables with markdown formatting:

code
## Feature Request Analysis

### Overview
| Metric | Count |
|--------|-------|
| Total Open | X |
| Roadmapped | X (Y%) |
| Unassigned | X (Y%) |

### High-Impact by Domain
[Table]

### Partner Volume
[Table]

### Multi-Partner Requests (Highest Leverage)
[Table]

### Unroadmapped High-Impact
[Table]

### Team Workload
[Table]

## Key Findings
1. ...
2. ...

## Recommendations
1. ...
2. ...

## Suggested Next Steps
- ...

Optional Deep Dives

If user asks, can drill into:

  • Specific domain: "Show me all [Domain] requests"
  • Specific partner: "What does [Partner] need?"
  • Specific team: "What's on [Team]'s plate?"
  • Trends: "What's been submitted recently?" (filter by Submission Date)
  • Stale items: "What's been sitting untriaged?" (old submission, no team, no roadmap)

Integration with Other Skills

  • After identifying high-value opportunities, suggest creating briefs using the Product Management Skill
  • For items already with briefs, can link to the briefs/ folder