Reviewing Progress

Overview

Assess progress toward an open-ended goal by reading iteration journals, summarizing achievements, identifying patterns, and suggesting next steps.

Core principle: Regular review reveals trends, validates strategy, and informs future direction.

When to Use

Use this skill when:

• User runs /review-progress command
• User asks "how are we doing?" or "what's our progress?"
• Need to assess if current strategy is working
• Deciding whether to pivot approach
• Preparing report on goal status

Can be used:

• Mid-iteration (doesn't end current iteration)
• Standalone (no active iteration required)
• Anytime user wants status check

Quick Reference

Step	Action	Tool/Agent
1. Locate goal	Find autonomy directory	Glob
2. Load summary	Read summary.md if exists	Read
3. Load recent iterations	Get last 3-5 iterations	Task (journal-reader)
4. Analyze progress	Identify trends, metrics, patterns	Manual analysis
5. Present report	Structured progress summary	Direct output

Process

Step 1: Locate Goal

Find the goal directory:

# Use Glob to find goal
pattern: "autonomy/*/goal.md"

If multiple goals found:

• List goals and ask user which to review
• If only one goal, proceed with that one

If no goal found:

"No autonomy goal found in this project. Use `/start-iteration` to begin tracking an open-ended goal."

Step 2: Load Summary (if exists)

Check for summary.md:

# Use Read to load summary
file: "autonomy/[goal-name]/summary.md"

If summary exists:

• Read full content
• Note which iterations it covers
• Use as foundation for report

If summary doesn't exist:

• Will rely on reading iterations directly

Step 3: Load Recent Iterations

Dispatch journal-reader to load recent context:

Task tool with subagent_type: "autonomy:journal-reader"
Prompt: "Read all iteration files for goal '[goal-name]' (or last 5-10 if many exist).
        Extract:
        - Progress timeline
        - Completed work by iteration
        - Persistent blockers
        - Metric trends
        - Skills most frequently used
        - Strategic pivots"
Model: haiku

Wait for agent response with structured findings.

Step 4: Analyze Progress

Synthesize information from summary and journal-reader:

Calculate metrics:

• If goal tracks metrics: Starting value → Current value → % change
• Iteration count: How many iterations completed?
• Time elapsed: First iteration date → Latest iteration date

Identify patterns:

• Which types of work have been most common?
• Are there recurring blockers?
• Has strategy evolved or stayed consistent?
• Which skills/workflows used most frequently?

Assess trajectory:

• Is progress accelerating, steady, or slowing?
• Are recent iterations more or less productive?
• Is current approach working?

Flag concerns:

• Blockers appearing in multiple iterations (persistent issues)
• Metrics stagnating or declining
• Lack of clear recent progress
• Strategy thrashing (frequent pivots without validation)

Step 5: Present Report

Generate comprehensive progress report:

# Progress Report: [Goal Name]

**Report Date:** [Today's date]
**Goal Status:** Active
**Iterations Completed:** [N]
**Time Elapsed:** [First date] - [Latest date] ([X days/weeks/months])

---

## Goal Statement
[From goal.md]

---

## Progress Overview

### Key Metrics
| Metric | Starting | Current | Change |
|--------|----------|---------|--------|
| [Metric 1] | [Value] | [Value] | [+X% or -Y%] |
| [Metric 2] | [Value] | [Value] | [+X% or -Y%] |

### Timeline of Major Work
- **Iteration 1:** [Summary of work]
- **Iteration 2:** [Summary of work]
- **Iteration N:** [Summary of work]

### Completed Initiatives
✅ [Initiative 1]: [Outcome]
✅ [Initiative 2]: [Outcome]

### In Progress
🚧 [Initiative 3]: [Current state]

---

## Current State

### What's Working Well
- [Positive pattern 1]
- [Positive pattern 2]

### Current Blockers
- **[Blocker 1]:** [First appeared iteration X, still unresolved]
- **[Blocker 2]:** [Description and impact]

### Open Questions
- [Question 1]
- [Question 2]

---

## Analysis

### Strategic Evolution
[How has the approach changed over time?]
- Iteration 1-3: [Initial strategy]
- Iteration 4-6: [Pivot or continuation]
- Current: [Where we are now]

### Skills & Methods Most Used
- **[Skill/workflow]:** Used in [X] iterations for [purpose]
- **[Skill/workflow]:** Used in [Y] iterations for [purpose]

### Effectiveness Assessment
[Is the current approach working?]
- **Strengths:** [What's effective]
- **Weaknesses:** [What's not working]
- **Opportunities:** [What could be explored next]

---

## Recommendations

### Immediate Next Steps
1. [Specific action based on current state]
2. [Specific action based on blockers]

### Strategic Considerations
- [Consider pivot if X]
- [Double down on Y because Z]
- [Investigate new approach for A]

### Health Check
[Overall assessment: Is goal on track? Should strategy change? Is this goal still valuable?]

---

**This is an open-ended goal - continuous iteration and optimization expected.**

Important Notes

Mid-Iteration Review

If reviewing during active iteration:

• Include current iteration work in report
• Note that current iteration is in progress
• Don't write journal entry (that's ending-an-iteration's job)

Standalone Review

If no iteration active:

• Report based solely on journal history
• Suggest starting new iteration if next steps are clear

Frequency Recommendations

Suggest reviewing progress:

• Every 5 iterations (natural checkpoint)
• When considering strategy pivot
• Before major decisions
• If feeling lost or uncertain

Honesty in Assessment

Be honest:

• If metrics aren't improving, say so
• If approach isn't working, acknowledge it
• If goal seems stalled, flag it
• Don't sugarcoat to avoid disappointing user

Constructive:

• Always include specific next steps
• Suggest concrete changes if needed
• Identify what IS working to preserve

Common Mistakes

Mistake	Reality
"Progress looks good" without data	NO. Use specific metrics and evidence.
"I'll review just the last iteration"	NO. Look at full history for patterns.
"No need to flag concerns"	NO. Honest assessment helps course-correct.
"I'll skip analysis and just list work done"	NO. Synthesis and insights are the value.
"Goal is open-ended so no progress metrics"	NO. Even open goals have measurable indicators.

After Reviewing

Once review is complete:

• User has clear picture of progress
• Can decide: continue current path, pivot strategy, or pause goal
• Next iteration can incorporate review insights
• May identify need to update goal.md with new success criteria