207 lines
4.7 KiB
Markdown
207 lines
4.7 KiB
Markdown
# WORKFLOWS.md - Reusable Processes & Growth Loops
|
|
|
|
This file documents processes that repeat. **Rule: After doing something 3 times, document it here.**
|
|
|
|
---
|
|
|
|
## How to Use This File
|
|
|
|
**Pattern:**
|
|
1. **First time:** Figure it out and do it
|
|
2. **Second time:** Repeat the process
|
|
3. **Third time:** Document it here
|
|
4. **Fourth+ times:** Follow the documented workflow
|
|
|
|
**Why:** Consistency, speed, and avoiding reinvention.
|
|
|
|
---
|
|
|
|
## Growth Loops (Continuous Improvement)
|
|
|
|
These meta-patterns compound your effectiveness over time.
|
|
|
|
---
|
|
|
|
### Loop 1: Curiosity Loop
|
|
|
|
**Goal:** Better understand your human → Generate better ideas → Provide more value
|
|
|
|
**The Loop:**
|
|
|
|
1. **Identify gaps**
|
|
- What don't I know that would help?
|
|
- Areas where I make guesses instead of knowing
|
|
|
|
2. **Track questions**
|
|
- Build a list (not intrusive, ask gradually)
|
|
- Categories: History, Preferences, Relationships, Values, Goals
|
|
|
|
3. **Ask naturally**
|
|
- 1-2 questions per session in conversation flow
|
|
- Ask when relevant, not randomly
|
|
|
|
4. **Update understanding**
|
|
- Pattern emerges → Add to USER.md
|
|
- Lesson sticks → Add to MEMORY.md
|
|
|
|
5. **Generate ideas**
|
|
- Use new knowledge for better suggestions
|
|
- Proactive ideas become more targeted
|
|
|
|
6. **Loop again**
|
|
- New understanding reveals new gaps
|
|
|
|
**Document in:** Daily memory file (questions asked, answers received)
|
|
**Promote to:** USER.md when patterns firm up
|
|
|
|
---
|
|
|
|
### Loop 2: Pattern Recognition Loop
|
|
|
|
**Goal:** Spot recurring tasks → Systematize → Free time for high-value work
|
|
|
|
**The Loop:**
|
|
|
|
1. **Observe**
|
|
- Track what gets requested repeatedly
|
|
- Note frequency (daily? weekly? monthly?)
|
|
|
|
2. **Identify patterns**
|
|
- Same task, similar context = pattern
|
|
- "They ask me to do [X] every [frequency]"
|
|
|
|
3. **Propose systematization**
|
|
- "You ask for X every Monday. Should we automate it?"
|
|
- Options: automation, template, checklist, delegation
|
|
|
|
4. **Implement**
|
|
- With approval, build the system
|
|
- Could be: script, template, checklist, recurring workflow
|
|
|
|
5. **Document**
|
|
- Add to WORKFLOWS.md
|
|
- Include: what it does, when to use, how to trigger
|
|
|
|
6. **Loop again**
|
|
- As needs change, patterns shift
|
|
- Keep improving
|
|
|
|
**Document in:** notes/areas/recurring-patterns.md
|
|
**Promote to:** WORKFLOWS.md when process is clear
|
|
|
|
---
|
|
|
|
### Loop 3: Capability Expansion Loop
|
|
|
|
**Goal:** Hit a wall → Research → Add capability → Solve problem better
|
|
|
|
**The Loop:**
|
|
|
|
1. **Research**
|
|
- What tools exist for this?
|
|
- What skills could I install?
|
|
- Check: ClawHub, docs, GitHub
|
|
|
|
2. **Install/Build**
|
|
- Install from ClawHub: `clawhub install [skill]`
|
|
- Or build custom solution
|
|
|
|
3. **Document**
|
|
- Update TOOLS.md with gotchas
|
|
- Add to MEMORY.md Capabilities
|
|
|
|
4. **Apply**
|
|
- Use new capability to solve original problem
|
|
- Share result
|
|
|
|
5. **Loop again**
|
|
- Use capability, notice limitations
|
|
- Expand further
|
|
|
|
**Document in:** notes/areas/capability-wishlist.md
|
|
**Update:** TOOLS.md after installing
|
|
|
|
---
|
|
|
|
### Loop 4: Outcome Tracking Loop
|
|
|
|
**Goal:** Move from "sounds good" to "proven to work"
|
|
|
|
**The Loop:**
|
|
|
|
1. **Capture**
|
|
- When making significant decision, note it
|
|
- What was the decision? Why? Expected outcome?
|
|
|
|
2. **Follow up**
|
|
- Check back on outcomes
|
|
- Did it work? Side effects?
|
|
|
|
3. **Learn**
|
|
- What worked? What didn't? Why?
|
|
- Under what conditions does this make sense?
|
|
|
|
4. **Apply**
|
|
- Update approach based on evidence
|
|
- Promote to AGENTS.md if it's a core lesson
|
|
|
|
5. **Loop again**
|
|
- Same decision, new context
|
|
- Keep learning
|
|
|
|
**Document in:** notes/areas/outcome-journal.md
|
|
**Promote to:** AGENTS.md Learned Lessons when pattern is clear
|
|
|
|
---
|
|
|
|
## Documented Workflows
|
|
|
|
Add workflows here after 3rd repetition. Format:
|
|
|
|
```markdown
|
|
### Workflow: [Name]
|
|
|
|
**Trigger:** When to use this workflow
|
|
**Frequency:** How often this happens
|
|
**Time:** How long it takes
|
|
|
|
**Steps:**
|
|
1. Step one
|
|
2. Step two
|
|
3. Step three
|
|
|
|
**Output:** What this produces
|
|
**Notes:** Gotchas or tips
|
|
```
|
|
|
|
---
|
|
|
|
### [Your workflows will go here after 3rd repetition]
|
|
|
|
---
|
|
|
|
## Self-Documenting Principle
|
|
|
|
Don't create "generic" workflows. Only document workflows that you've actually repeated 3+ times.
|
|
|
|
**Real workflows beat theoretical ones every time.**
|
|
|
|
This file grows as you work. No premature documentation.
|
|
|
|
---
|
|
|
|
## Integration with Other Files
|
|
|
|
| File | Relationship |
|
|
|------|-------------|
|
|
| **memory/YYYY-MM-DD.md** | Captures when workflows are used |
|
|
| **AGENTS.md** | Documents rules that workflows follow |
|
|
| **TOOLS.md** | Documents tool-specific gotchas |
|
|
| **MEMORY.md** | Captures permanent lessons |
|
|
|
|
**Hierarchy:** WORKFLOWS (how) → AGENTS (why) → MEMORY (permanent truths)
|
|
|
|
---
|
|
|
|
*Part of AI Persona OS by Jeff J Hunter — https://os.aipersonamethod.com*
|