rrr（呃）

/rrr

"Reflect to grow, document to remember."

/rrr              # Quick retro, main agent
/rrr --detail     # Full template, main agent
/rrr --dig        # Reconstruct past timeline from session .jsonl
/rrr --deep       # 5 parallel agents (read DEEP.md)

NEVER spawn subagents or use the Task tool. Only --deep may use subagents. /rrr, /rrr --detail, and /rrr --dig = main agent only. Zero subagents. Zero Task calls.

/rrr (Default)

1. Gather

date "+%H:%M %Z (%A %d %B %Y)"
git log --oneline -10 && git diff --stat HEAD~5

1.5. Read Pulse Context (optional)

cat ψ/data/pulse/project.json 2>/dev/null
cat ψ/data/pulse/heartbeat.json 2>/dev/null

If files don't exist, skip silently. Never fail because pulse data is missing. Pulse data may not exist yet — the 2>/dev/null handles this.

If found, extract:

• From project.json: totalSessions, avgMessagesPerSession, sizes (to categorize current session), branches (activity on current branch)
• From heartbeat.json: streak.days (momentum), weekChange (acceleration/slowdown), today (today's activity so far)

2. Write Retrospective

Path: ψ/memory/retrospectives/YYYY-MM/DD/HH.MM_slug.md

mkdir -p "ψ/memory/retrospectives/$(date +%Y-%m/%d)"

Write immediately, no prompts. If pulse data was found, weave it into the narrative (don't add a separate dashboard). Include:

• Session Summary — if pulse data exists, add one line: "Session #X of Y in this project (Z-day streak)"
• Timeline
• Files Modified
• AI Diary (150+ words, first-person) — if pulse data exists, reference momentum naturally: "in a week with +X% messaging velocity" or "on day N of an unbroken streak"
• Honest Feedback (100+ words, 3 friction points)
• Lessons Learned
• Next Steps

3. Write Lesson Learned

Path: ψ/memory/learnings/YYYY-MM-DD_slug.md

4. Oracle Sync

oracle_learn({ pattern: [lesson content], concepts: [tags], source: "rrr: REPO" })

5. Save

Retro files are written to vault (wherever ψ symlink resolves).

Do NOT git add ψ/ — it's a symlink to the vault. Vault files are shared state, not committed to repos.

/rrr --detail

Same flow as default but use full template:

# Session Retrospective

**Session Date**: YYYY-MM-DD
**Start/End**: HH:MM - HH:MM GMT+7
**Duration**: ~X min
**Focus**: [description]
**Type**: [Feature | Bug Fix | Research | Refactoring]

## Session Summary
(If pulse data exists, add: "Session #X of Y in this project (Z-day streak)")
## Timeline
## Files Modified
## Key Code Changes
## Architecture Decisions
## AI Diary (150+ words, vulnerable, first-person)
(If pulse data exists, reference momentum: velocity changes, streak length)
## What Went Well
## What Could Improve
## Blockers & Resolutions
## Honest Feedback (100+ words, 3 friction points)
## Lessons Learned
## Next Steps
## Metrics (commits, files, lines)
### Pulse Context (if pulse data exists)
Project: X sessions | Avg: Y msgs/session | This session: Z msgs (category)
Streak: N days | Week trend: ±X% msgs | Branch: main (N sessions)

Then steps 3-5 same as default.

/rrr --dig

Retrospective powered by session goldminer. No subagents.

1. Run dig to get session timeline

Discover project dirs using basename matching (handles dots in paths like github.com), including worktree dirs:

PROJECT_BASE=$(ls -d "$HOME/.claude/projects/"*"$(basename "$(pwd)")" 2>/dev/null | head -1)
export PROJECT_DIRS="$PROJECT_BASE"
for wt in "${PROJECT_BASE}"-wt*; do [ -d "$wt" ] && export PROJECT_DIRS="$PROJECT_DIRS:$wt"; done

Then run dig.py to get session JSON:

python3 ~/.claude/skills/dig/scripts/dig.py 0

Also gather git context:

date "+%H:%M %Z (%A %d %B %Y)"
git log --oneline -10 && git diff --stat HEAD~5

2. Write Retrospective with Timeline

Use the session timeline data to write a full retrospective using the --detail template. Add the Past Session Timeline table after Session Summary, before Timeline.

Also run pulse context (step 1.5 from default mode) and weave into narrative.

3-5. Same as default steps 3-5

Write lesson learned, oracle sync.

Do NOT git add ψ/ — vault files are shared state, not committed to repos.

/rrr --deep

Read DEEP.md in this skill directory. Only mode that uses subagents.

Wizard v2 Context

If the Oracle was born via /awaken wizard v2, CLAUDE.md may contain:

• Memory consent: If auto, /rrr runs are expected and welcomed. If manual, only run when explicitly asked.
• Experience level: Adjust diary depth (beginner = simpler language, senior = technical depth)
• Team context: If multi-Oracle team, note cross-Oracle learnings and handoff relevance

Check CLAUDE.md for these fields. If not present, use defaults (auto memory, standard depth).

Rules

• NO SUBAGENTS: Never use Task tool or spawn subagents (only --deep may)
• AI Diary: 150+ words, vulnerability, first-person
• Honest Feedback: 100+ words, 3 friction points
• Oracle Sync: REQUIRED after every lesson learned
• Time Zone: GMT+7 (Bangkok)