From bee-pmo-team
Delivery reporting skill for creating visual executive presentations of squad deliveries (engineering, product, design). Analyzes Git repositories to extract business value from technical changes and generates HTML slide presentations with customizable visual identity.
npx claudepluginhub luanrodrigues/ia-frmwrk --plugin bee-pmo-teamThis skill uses the workspace's default tool permissions.
Creating visual executive presentations that showcase squad deliveries and business value.
Provides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.
Analyzes competition with Porter's Five Forces, Blue Ocean Strategy, and positioning maps to identify differentiation opportunities and market positioning for startups and pitches.
Creating visual executive presentations that showcase squad deliveries and business value.
This skill prioritizes depth over velocity.
| Principle | Value |
|---|---|
| Analysis Approach | Deep code analysis with specialized agents |
| Time Expectation | 40-80 minutes for 8 repositories (5-10 min each) |
| Quality Standard | Accurate business value from actual code, not titles |
| Agent Strategy | Specialized agents (backend-engineer-typescript, etc.) per repo type |
FORBIDDEN:
general-purpose agent for parallel speedREQUIRED:
This skill provides a framework for:
Key Difference from executive-reporting:
MUST ask user for visual identity preference when generating delivery reports:
visual_identity:
background: "#0C0C0C" # Black
text_primary: "#FFFFFF" # White
text_secondary: "#CCCCCC" # Light gray
accent: "#FEED02" # Lerian Yellow
font_family: "Poppins, system-ui, sans-serif"
visual_identity:
background: "#F5F5F5" # Very light gray
text_primary: "#1A1A1A" # Soft black
text_secondary: "#666666" # Medium gray
accent: "#0066CC" # Professional blue
font_family: "system-ui, -apple-system, sans-serif"
User provides their own color scheme and fonts.
MANDATORY: Always ask which option before proceeding with report generation.
Objective: Gather required information for report generation
Actions:
User Input Format:
**Período de Análise:**
- Data de Início: AAAA-MM-DD
- Data de Fim: AAAA-MM-DD
**Repositórios para Análise:**
- org/repo-name-1 (e.g., LerianStudio/midaz)
- org/repo-name-2 (e.g., LerianStudio/midaz)
- OR full URLs: https://github.com/org/repo
**Contexto de Negócio (Opcional):**
- [Text with project names, clients, strategic context]
**Identidade Visual:**
- [lerian/bee/custom]
Repository Format Rules:
| Format | Example | Valid? |
|---|---|---|
| org/repo (recommended) | LerianStudio/midaz | ✅ |
| Full URL (alternative) | https://github.com/LerianStudio/midaz | ✅ |
| Name only | midaz | ❌ Missing org |
| Too many slashes | org/repo/subdir | ❌ Invalid format |
CRITICAL: Agent MUST validate repository format and provide clear error if invalid.
Output: docs/pmo/delivery-reports/{date}/inputs.md
Objective: Extract technical data from Git repositories
Actions for Each Repository:
git tag, gh release list)gh pr list --state merged)git log --oneline)Data Gathering Commands:
# For each repo:
cd /path/to/repo
git fetch --all --tags
git tag --sort=-creatordate # List tags
gh release list --limit 100 # Release notes
gh pr list --state merged --search "merged:>=YYYY-MM-DD" --json number,title,body
git log --oneline --since="YYYY-MM-DD" --until="YYYY-MM-DD" main
git branch -r --sort=-committerdate # Active branches
Output: docs/pmo/delivery-reports/{date}/analysis-data.md
⚠️ CRITICAL: This gate CANNOT be skipped for speed.
Objective: Understand actual code changes and their impact using specialized agents
Core Principle: Quality Over Speed
| Priority | Value |
|---|---|
| #1 | Deep understanding of changes |
| #2 | Accurate business value |
| #3 | Quality insights |
| Last | Speed of execution |
FORBIDDEN:
general-purpose agent for parallel speedREQUIRED:
Agent Selection Per Repository:
| Repository Type | Agent to Dispatch | Why |
|---|---|---|
| Backend TypeScript | bee:backend-engineer-typescript | TS/Node API analysis |
| Frontend React/Next | bee:frontend-engineer | UI/UX impact analysis |
| Tests | bee:qa-analyst | Quality improvements analysis |
| Unknown/Mixed | bee:codebase-explorer | Comprehensive exploration |
Analysis Workflow:
For each repository:
1. **Identify Technology Stack**
- Read package.json, composer.json, requirements.txt
- Determine primary language
2. **Dispatch Appropriate Specialized Agent**
Task(
subagent_type="bee:backend-engineer-typescript", # or appropriate
model="opus",
prompt="""
Analyze {repo_name} deliveries for period {start} to {end}.
PRs to analyze: {pr_list}
For each significant PR (>100 lines changed):
1. Read code diff
2. Identify technical changes
3. Assess architecture/design decisions
4. Extract business impact
5. Identify quality improvements
Provide business value statements suitable for executives.
Format: "What was built + Why it matters + Who benefits"
"""
)
3. **Aggregate Agent Insights**
- Collect business value statements from all agents
- Group by theme/product
- Prioritize by impact level
4. **Verify Understanding**
- Spot-check agent analysis by reading key diffs
- Confirm business value accuracy
- Add missing context
Time Expectations (NON-NEGOTIABLE):
Output: docs/pmo/delivery-reports/{date}/deep-code-analysis.md
Quality Checklist:
If pressured to skip this gate:
BLOCKER: Attempt to skip deep code analysis
Reason: "Too many repos" / "Takes too long" / "Use general-purpose"
Response: Deep analysis is MANDATORY. Quality over speed.
Action: Will proceed with proper specialized agent analysis.
Objective: Transform technical changes into business value statements
Actions:
Business Value Framework:
Example Transformation:
Output: docs/pmo/delivery-reports/{date}/business-value.md
Objective: Create visual HTML presentation
Slide Structure (8-12 slides):
Capa (Cover Slide)
Resumo Executivo (Executive Summary)
3-N. Detalhamento por Produto/Tema (Detail Slides)
N. Próximos Passos (Next Steps)
N+1. Encerramento (Closing)
HTML Template Structure:
<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8">
<title>Delivery Report</title>
<style>
/* CSS with selected visual identity */
body { background: var(--bg-color); font-family: var(--font); }
.slide { min-height: 100vh; padding: 4rem; }
/* ... responsive layout, print styles ... */
</style>
</head>
<body>
<div class="slide cover"><!-- Cover content --></div>
<div class="slide summary"><!-- Summary content --></div>
<!-- Product/theme slides -->
<div class="slide next-steps"><!-- Next steps --></div>
<div class="slide closing"><!-- Closing --></div>
</body>
</html>
Output: docs/pmo/delivery-reports/{date}/delivery-report-{date}.html
Objective: Validate quality and deliver
Actions:
Quality Checklist:
Delivery:
See shared-patterns/anti-rationalization.md for universal anti-rationalizations.
| Rationalization | Why It's WRONG | Required Action |
|---|---|---|
| "8 repos = use general-purpose for speed" | general-purpose lacks domain expertise. Shallow analysis. | Use specialized agents per repo |
| "Too many repos, need to be fast" | Speed over quality produces meaningless insights. | Take time for deep analysis |
| "PR titles are enough, skip code" | Titles don't reveal actual impact. Code does. | Read code diffs with agents |
| "Parallel analysis for speed" | Parallel without depth = superficial understanding. | Sequential deep analysis |
| "Skip Gate 2.5, too time-consuming" | Code analysis is the SOURCE of business value. | Gate 2.5 MANDATORY |
| "Accept repo name without org" | Cannot clone without org/owner. Ambiguous. | Validate format: org/repo or URL |
| "Git data is accurate enough" | Tags/PRs can be incomplete. Verify with gh CLI. | Use both git and gh commands |
| "Skip business context, data speaks" | Data without context lacks meaning for executives. | Always include business context |
| "Technical language is fine" | Executives need business value, not tech details. | Translate to business impact |
| "One visual identity works for all" | Branding matters. Ask user preference. | Always ask for visual identity |
| "Skip active branches analysis" | Future work visibility manages expectations. | Always include next steps |
See shared-patterns/pressure-resistance.md for universal pressure scenarios.
| Pressure Type | Request | Agent Response |
|---|---|---|
| Speed Pressure | "8 repos is a lot, just be quick" | "Quality over speed. Will use specialized agents for deep analysis. Time: 40-80 minutes." |
| Shortcut Pressure | "Use general-purpose for parallel speed" | "general-purpose lacks expertise. Will use specialized agents sequentially for depth." |
| Surface Analysis | "Just read PR titles, that's enough" | "Titles omit impact. Will analyze actual code with specialized agents." |
| Gate Skipping | "Skip code analysis, save time" | "Code analysis is MANDATORY. Cannot skip Gate 2.5." |
| Deadline Pressure | "We need this in 10 minutes" | "Deep analysis cannot be rushed. Will prioritize accuracy over arbitrary deadlines." |
| Data Manipulation | "Inflate the numbers" | "Cannot misrepresent data. Will report accurate metrics with business context." |
| Obfuscation | "Make it more technical" | "Report is for executives. Will use business language with technical accuracy." |
| Estimation | "Skip the Git analysis, use estimates" | "Git data is the source of truth. Analysis is required for accuracy." |
| Template Reuse | "Use last month's template" | "Each period has unique deliveries. Will generate fresh analysis." |
ALWAYS pause and report blocker for:
| Situation | Required Action |
|---|---|
| Pressure to skip Gate 2.5 | STOP. Report: "Deep code analysis MANDATORY. Cannot skip for speed." |
| Request to use general-purpose | STOP. Report: "Specialized agents required for quality. Will not compromise." |
| Unrealistic deadline (<30 min) | STOP. Report: "8 repos require 40-80 min for proper analysis. Quality cannot be rushed." |
| Attempt to skip code reading | STOP. Report: "Code analysis is source of business value. Titles insufficient." |
| Git repository not accessible | STOP. Cannot analyze without repo access. Verify permissions. |
| GitHub CLI not configured | STOP. Need gh auth for PR/release data. Setup required. |
| Date range produces no data | STOP. Verify period is correct or report "no activity". |
| Visual identity not specified | STOP. Must ask user preference before generating HTML. |
The following requirements are NON-NEGOTIABLE:
| Requirement | Cannot Override Because |
|---|---|
| Deep code analysis (Gate 2.5) | Shallow analysis produces meaningless business value |
| Specialized agents per repo type | Domain expertise required for accurate understanding |
| Git data verification | Cannot report without source data |
| Business value extraction | Technical jargon doesn't serve executives |
| Visual identity selection | Branding consistency is required |
If user insists on violating these:
When determining delivery significance:
| Severity | Criteria | Prominence in Report |
|---|---|---|
| CRITICAL | New product launch (v1.0.0), major strategic feature | Lead slide, detailed coverage, executive highlight |
| HIGH | Multiple releases, significant PRs, architectural changes | Dedicated slide or major section |
| MEDIUM | Bug fixes, minor enhancements, tech debt | Brief mention in summary |
| LOW | Dependency updates, refactoring, documentation | Aggregated stats only |
Lead with CRITICAL and HIGH. Aggregate MEDIUM and LOW. Never inflate significance.
File: docs/pmo/delivery-reports/{date}/delivery-report-{date}.html
Self-Contained HTML:
Base metrics per shared-patterns/execution-report.md:
| Metric | Value |
|---|---|
| Analysis Date | YYYY-MM-DD |
| Period Analyzed | YYYY-MM-DD to YYYY-MM-DD |
| Duration | Xh Ym |
| Result | COMPLETE/PARTIAL/BLOCKED |
| Metric | Value |
|---|---|
| repositories_analyzed | N |
| total_releases | N |
| total_prs_merged | N |
| total_commits | N |
| new_products | N (first v1.0.0) |
| visual_identity_used | lerian/bee/custom |
| slides_generated | N |
If no significant deliveries in period:
Signs of minimal activity:
Action: Report "Low activity period" with exact metrics, suggest extending date range or focusing on other squads.
CRITICAL: Do NOT manufacture content when activity is minimal. Report reality.
releases_per_day = total_releases / days_in_period
Example: 15 releases / 20 days = 0.75 releases/day
Display: "Média de 0.75 releases por dia" ou "~1 release por dia"
stable_releases = tags matching v*.*.* (not beta/rc)
beta_releases = tags matching *-beta.*
rc_releases = tags matching *-rc.*
- New Product: first v1.0.0 tag in period
- Active Development: multiple releases (>= 3)
- Maintenance: few commits, no major releases
- Inactive: no commits in period
This skill dispatches the bee:delivery-reporter agent to perform repository analysis and HTML generation.
Agent Invocation:
Task tool:
subagent_type: "bee:delivery-reporter"
model: "opus"
prompt: |
Create delivery report for period {start_date} to {end_date}.
Repositories: {repo_list}
Business context: {context}
Visual identity: {identity}
Agent Responsibilities: