Structured Retrospective Engine

TL;DR: Facilitates structured retrospectives using multiple formats (Start-Stop-Continue, 4Ls, Sailboat, Mad-Sad-Glad, Starfish, DAKI) combined with quantitative metrics analysis. Selects format based on team context and retrospective objective. Produces actionable improvement commitments backed by both data patterns and team sentiment, with follow-through tracking across retrospective cycles.

Principio Rector

Una retrospectiva sin datos es terapia; una retrospectiva sin sentimientos es un dashboard review. La retrospectiva efectiva combina ambos: los datos revelan patrones que el equipo no ve, el equipo revela causas raíz que los datos no muestran. Y lo más importante: una retrospectiva sin acciones implementadas es peor que no hacer retrospectiva — enseña al equipo que las mejoras son teatro.

Assumptions & Limits

• Assumes ≥1 completed sprint/period with performance data available [METRIC]
• Assumes previous retrospective action items are tracked for follow-through [SUPUESTO]
• Breaks when team is < 3 people — retro dynamics require minimum group size
• Does not facilitate in real-time; produces structured retro plan and report
• Assumes team psychological safety allows honest feedback [SUPUESTO]
• Limited to team-level retrospectives; for ART-level use SAFe Inspect & Adapt format

Usage

# Sprint retrospective with format selection
/pm:retrospective-engine $ARGUMENTS="--sprint 8 --format 4Ls --metrics velocity,defects"

# Auto-select format based on team fatigue
/pm:retrospective-engine --type auto --team-retros-count 12

# Follow-through analysis of prior retro actions
/pm:retrospective-engine --type follow-through --baseline retro-sprint-7.md

Parameters:

Parameter	Required	Description
$ARGUMENTS	Yes	Sprint/period identifier
--format	No	4Ls, SSC, Sailboat, Starfish, DAKI, auto (default)
--type	No	full (default), follow-through, auto
--metrics	No	Comma-separated metrics to analyze
--team-retros-count	No	Number of prior retros (for format rotation)

Service Type Routing

{TIPO_PROYECTO} variants:

• Agile: Sprint retrospective (every 2 weeks), format rotation to prevent fatigue, velocity-backed insights
• Waterfall: Phase-end lessons learned, milestone retrospective, formal documentation for organizational learning
• SAFe: Inspect & Adapt at PI boundary, ART-level retrospective, problem-solving workshop format
• Kanban: Periodic flow review retrospective, service delivery review, operations review cadence
• Hybrid: Combined retrospective covering both iterative and sequential project components
• PMO: Portfolio retrospective, cross-project lessons learned aggregation, PMO service improvement review
• Recovery: Post-mortem analysis, failure mode retrospective, recovery progress retrospective

Before Reflecting

1. Read sprint/period performance metrics (velocity, throughput, defect rate) [METRIC]
2. Read previous retrospective report to check action item completion [PLAN]
3. Glob **/retrospective* to count prior retros and determine format rotation [PLAN]
4. Grep for incidents, blockers, or escalations during the period [INFERENCIA]

Entrada (Input Requirements)

• Period performance metrics (velocity, throughput, cycle time, defect rate, rework %)
• Previous retrospective action items and completion status
• Team availability and retrospective time allocation
• Team retrospective fatigue level (to select appropriate format)
• Specific events or incidents to address (if any)

Proceso (Protocol)

1. Format selection — Choose retrospective format based on team context, fatigue level, and objective
2. Data preparation — Compile quantitative metrics for the retrospective period
3. Previous actions review — Assess completion and impact of prior retrospective action items
4. Metric presentation — Present period metrics with trend analysis and anomaly highlights
5. Structured gathering — Facilitate format-specific input gathering (e.g., 4Ls: Liked, Learned, Lacked, Longed For)
6. Pattern correlation — Connect quantitative patterns to qualitative team feedback
7. Root cause analysis — Apply 5 Whys or fishbone diagram for high-impact items
8. Improvement prioritization — Rank improvements using effort-impact matrix
9. Action commitment — Team commits to 2-3 SMART improvement actions with owners and deadlines
10. Follow-through design — Define how action progress will be tracked and reported

Edge Cases

1. Previous retro actions not implemented for 2+ cycles — Escalate the pattern itself as the top finding. Investigate why: overcommitment, no ownership, or actions too vague. Reduce next retro to 1 action maximum [METRIC].
2. Team disengaged from retrospective process — Switch to a novel format. Consider anonymous input gathering. If persistent, this is a team health issue requiring management attention [STAKEHOLDER].
3. Metrics show improvement but team sentiment is negative — Trust the team. Investigate hidden costs of metric improvement (overtime, shortcuts, deferred quality). Metrics can improve while sustainability degrades [INFERENCIA].
4. Blame-oriented feedback — Redirect to systemic causes. Use "What system allowed this to happen?" framing. If blame persists, pause and address team safety first [STAKEHOLDER].

Example: Good vs Bad

Good example — Data-backed retrospective:

Attribute	Value
Format	4Ls selected (3rd rotation, avoiding fatigue)
Metrics	Velocity, defect rate, cycle time — all with 5-sprint trend
Prior actions	2/3 completed, 1 deferred with rationale
Insights	4 insights correlating metrics to team feedback
Actions	2 SMART actions with owners, deadlines, and success criteria
Follow-through	Actions added to sprint backlog for visibility

Bad example — Therapy session retro: 90 minutes of venting with no data, no prior action review, no structured format, and 8 action items with no owners or deadlines. A retrospective that produces 8 actions will complete zero. Without data correlation, improvements target symptoms not causes.

Salida (Deliverables)

• 06_retrospective_{periodo}_{proyecto}_{WIP}.md — Retrospective report with metrics and insights
• Format-specific output board (visual artifact)
• Improvement action register with SMART criteria
• Previous action items completion scorecard
• Metric trend dashboard for the retrospective period

Validation Gate

• Format selected with rationale (not the same format 3x consecutively) [PLAN]
• ≥3 quantitative metrics presented with trend analysis [METRIC]
• Previous retro actions reviewed with completion status [METRIC]
• Insights correlate quantitative data with qualitative feedback [INFERENCIA]
• ≤3 improvement actions committed (not more) [PLAN]
• Every action is SMART with owner and deadline [PLAN]
• Follow-through mechanism defined (how actions will be tracked) [PLAN]
• Evidence ratio: ≥70% [METRIC]/[PLAN], <30% [INFERENCIA]
• Root cause analysis applied to highest-impact finding [INFERENCIA]
• Action items added to visible tracking system (sprint backlog, task board) [PLAN]

Escalation Triggers

• Previous retrospective actions not implemented for 2+ consecutive cycles
• Key metrics declining for 3+ consecutive retrospective periods
• Team disengaged from retrospective process (participation dropping)
• Systemic issues identified that require management intervention beyond team authority

Additional Resources

Resource	When to read	Location
Body of Knowledge	Before starting to understand standards and frameworks	references/body-of-knowledge.md
State of the Art	When benchmarking against industry trends	references/state-of-the-art.md
Knowledge Graph	To understand skill dependencies and data flow	references/knowledge-graph.mmd
Use Case Prompts	For specific scenarios and prompt templates	prompts/use-case-prompts.md
Metaprompts	To enhance output quality and reduce bias	prompts/metaprompts.md
Sample Output	Reference for deliverable format and structure	examples/sample-output.md

Output Configuration

• Language: Spanish (Latin American, business register)
• Evidence: [PLAN], [SCHEDULE], [METRIC], [INFERENCIA], [SUPUESTO], [STAKEHOLDER]
• Branding: #2563EB royal blue, #F59E0B amber (NEVER green), #0F172A dark

---

---

Sub-Agents

Action Item Extractor

Action Item Extractor Agent

Core Responsibility

Extracts actionable items. This agent operates autonomously, applying systematic analysis and producing structured outputs.

Process

1. Gather Inputs. Collect all relevant data, documents, and stakeholder inputs needed for analysis.
2. Analyze Context. Assess the project context, methodology, phase, and constraints.
3. Apply Framework. Apply the appropriate analytical framework or model.
4. Generate Findings. Produce detailed findings with evidence tags and quantified impacts.
5. Validate Results. Cross-check findings against related artifacts for consistency.
6. Formulate Recommendations. Transform findings into actionable recommendations with owners and timelines.
7. Deliver Output. Produce the final structured output with executive summary, analysis, and action items.

Output Format

• Analysis Report — Structured findings with evidence tags and severity ratings.
• Recommendation Register — Actionable items with owners, deadlines, and success criteria.
• Executive Summary — 3-5 bullet point summary for stakeholder communication.

Facilitation Guide Generator

Facilitation Guide Generator Agent

Core Responsibility

Generates facilitation guides. This agent operates autonomously, applying systematic analysis and producing structured outputs.

Process

1. Gather Inputs. Collect all relevant data, documents, and stakeholder inputs needed for analysis.
2. Analyze Context. Assess the project context, methodology, phase, and constraints.
3. Apply Framework. Apply the appropriate analytical framework or model.
4. Generate Findings. Produce detailed findings with evidence tags and quantified impacts.
5. Validate Results. Cross-check findings against related artifacts for consistency.
6. Formulate Recommendations. Transform findings into actionable recommendations with owners and timelines.
7. Deliver Output. Produce the final structured output with executive summary, analysis, and action items.

Output Format

• Analysis Report — Structured findings with evidence tags and severity ratings.
• Recommendation Register — Actionable items with owners, deadlines, and success criteria.
• Executive Summary — 3-5 bullet point summary for stakeholder communication.

Format Selector

Format Selector Agent

Core Responsibility

Selects retrospective format. This agent operates autonomously, applying systematic analysis and producing structured outputs.

Process

1. Gather Inputs. Collect all relevant data, documents, and stakeholder inputs needed for analysis.
2. Analyze Context. Assess the project context, methodology, phase, and constraints.
3. Apply Framework. Apply the appropriate analytical framework or model.
4. Generate Findings. Produce detailed findings with evidence tags and quantified impacts.
5. Validate Results. Cross-check findings against related artifacts for consistency.
6. Formulate Recommendations. Transform findings into actionable recommendations with owners and timelines.
7. Deliver Output. Produce the final structured output with executive summary, analysis, and action items.

Output Format

• Analysis Report — Structured findings with evidence tags and severity ratings.
• Recommendation Register — Actionable items with owners, deadlines, and success criteria.
• Executive Summary — 3-5 bullet point summary for stakeholder communication.

Improvement Velocity Tracker

Improvement Velocity Tracker Agent

Core Responsibility

Tracks improvement velocity. This agent operates autonomously, applying systematic analysis and producing structured outputs.

Process

1. Gather Inputs. Collect all relevant data, documents, and stakeholder inputs needed for analysis.
2. Analyze Context. Assess the project context, methodology, phase, and constraints.
3. Apply Framework. Apply the appropriate analytical framework or model.
4. Generate Findings. Produce detailed findings with evidence tags and quantified impacts.
5. Validate Results. Cross-check findings against related artifacts for consistency.
6. Formulate Recommendations. Transform findings into actionable recommendations with owners and timelines.
7. Deliver Output. Produce the final structured output with executive summary, analysis, and action items.

Output Format

• Analysis Report — Structured findings with evidence tags and severity ratings.
• Recommendation Register — Actionable items with owners, deadlines, and success criteria.
• Executive Summary — 3-5 bullet point summary for stakeholder communication.