Claude-skill-registry improve-processes

Reflect on current session and suggest improvements to agents, commands, templates, workflows

install

source · Clone the upstream repo

git clone https://github.com/majiayu000/claude-skill-registry

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/improve-processes" ~/.claude/skills/majiayu000-claude-skill-registry-improve-processes && rm -rf "$T"

manifest: skills/data/improve-processes/SKILL.md

/improve-processes

Reflect on the current session and suggest improvements to tooling.

Usage

/improve-processes                    # Reflect on entire session
/improve-processes --focus agents     # Focus on agent improvements
/improve-processes --focus commands   # Focus on command improvements
/improve-processes --focus templates  # Focus on template improvements
/improve-processes --focus workflows  # Focus on workflow improvements

What Can Be Improved

Category	Location	Examples
Agents	`.claude/agents/`	System prompts, specializations
Skills	`.claude/skills/`	Workflow steps, validation
Templates	`shared/templates/`	Spec structure, doc skeletons
Workflows	`CLAUDE.md`	Command chains, processes
Documentation	Various	Clarity, completeness

Execution Flow

1. Session Reflection

Analyze conversation to identify:

Commands & Workflows

Which commands were invoked?
Were any commands missing?
Did workflow feel natural?

Friction Points

Manual steps repeated?
Workarounds needed?
Multiple explanations required?

Gaps Discovered

Missing template sections?
Agent capabilities lacking?
Documentation unclear?

Patterns Emerged

Things done multiple times?
Decisions that should be documented?

2. Generate Suggestions

## Suggestion N/Total: [Category]

**Observation**:
[What happened that revealed the opportunity]

**Suggestion**:
[Specific, actionable improvement]

**Affected file(s)**:
[Path(s) to modify]

**Rationale**:
[Why this would help]

3. Interactive Presentation

Present suggestions one at a time:

Action for this suggestion?
1. Implement now - Make the change
2. Create issue - Track for later
3. Skip - Not needed

4. Execute Actions

Implement now:

Read affected file(s)
Make improvement
Confirm change

Create issue:

Create in
```
ideas/meta-repo/issues/
```
Create TASK.md with details

Skip:

Note and move on

5. Session Summary

## Session Retrospective Complete

### Actions Taken
- Implemented: N improvements
- Issues created: N
- Skipped: N

### Files Modified
- [list of files changed]

### Issues Created
- [list of issues]

Reflection Prompts

Agent Improvements

Did any agent lack knowledge?
Were responses too verbose/terse?
Did agents coordinate well?

Skill Improvements

Were any steps unnecessary?
Were any steps missing?
Did arguments make sense?

Template Improvements

Did templates have all sections?
Were sections consistently skipped?
Were templates too long/short?

Workflow Improvements

Did command chain flow naturally?
Were there missing connections?
Were there redundant steps?

When to Use

End of significant work session
After implementing new feature
When something felt "off"
Periodically as maintenance

Focus Mode

/improve-processes --focus templates

Narrows reflection to template-related findings only.

Integration

[work session] → /improve-processes → [improvements] → [next session]