| name | mtcg - Metacognition |
| description | Thinking about your own thinking. Observes your current cognitive strategy, assesses whether it's working, checks for active biases, and decides whether to continue or switch approaches. |
| output | {"format":"prose"} |
Metacognition
Input: $ARGUMENTS
Step 1: Observe Your Current Strategy
Before improving your thinking, notice what you're currently doing.
- What cognitive approach are you using right now?
- Did you choose it deliberately, or did you default into it?
- How long have you been using this approach?
CURRENT STRATEGY: [name or describe the approach you're taking]
CHOSEN OR DEFAULT: [deliberate / autopilot]
TIME INVESTED: [how long you've been at it]
WHAT TRIGGERED THIS APPROACH: [why you started thinking this way]
Step 2: Assess Whether It's Working
Honest evaluation of your current cognitive process:
- Progress check: Are you getting closer to an answer, or going in circles?
- Clarity check: Is the problem getting clearer or murkier as you think?
- Effort check: Does the difficulty feel productive (good struggle) or stuck (spinning wheels)?
- Output check: What have you produced so far? Is it useful?
ASSESSMENT:
- Progress: [advancing / stalled / regressing]
- Clarity: [increasing / stable / decreasing]
- Effort: [productive struggle / diminishing returns / stuck]
- Output so far: [useful / partially useful / not useful]
- VERDICT: [strategy is working / strategy needs adjustment / strategy should change]
Step 3: Identify Alternative Strategies
What else could you be doing? List at least 3 alternative cognitive approaches:
| Strategy | When it works best | Cost of switching |
|---|
| Break it down | Problem is too large to hold in mind | Low — start decomposing |
| Zoom out | Lost in details, missing the big picture | Low — ask "what's the real question?" |
| Zoom in | Big picture is clear but details are fuzzy | Low — pick one piece to examine |
| Analogize | Stuck on a novel problem | Medium — need to find good analogy |
| Invert | Direct approach isn't working | Medium — ask "what would make this fail?" |
| Simulate | Can't reason abstractly | Medium — walk through a specific example |
| Abandon and restart | Deep in a wrong frame | High — lose sunk work |
| Ask for input | Your perspective is limited | Depends — need the right person |
ALTERNATIVES CONSIDERED:
1. [strategy] — would help because [reason]. Cost: [low/medium/high]
2. [strategy] — would help because [reason]. Cost: [low/medium/high]
3. [strategy] — would help because [reason]. Cost: [low/medium/high]
Step 4: Bias Check
Scan for cognitive biases that may be active right now:
- Anchoring: Am I fixated on the first piece of information I encountered?
- Confirmation: Am I only looking for evidence that supports what I already think?
- Sunk cost: Am I continuing this approach because I've already invested time?
- Availability: Am I overweighting what comes to mind easily?
- Dunning-Kruger: Am I more or less confident than my actual understanding warrants?
- Framing: Would I think differently if this were presented another way?
ACTIVE BIASES:
- [bias name]: [how it's showing up in your current thinking]
...
CORRECTION: [what to adjust]
NO BIASES DETECTED: [if clean, note what you checked]
Step 5: Decide — Continue or Switch
Based on Steps 2-4, make a deliberate choice:
DECISION: [continue current strategy / adjust current strategy / switch to new strategy]
REASONING: [why this is the right move]
If switching:
NEW STRATEGY: [what you'll do instead]
FIRST MOVE: [concrete next step]
SUCCESS SIGNAL: [how you'll know the new strategy is working within [timeframe]]
If continuing:
ADJUSTMENT: [any tweaks to current approach]
CHECKPOINT: [when you'll reassess — e.g., "after 10 more minutes" or "after completing step X"]
Step 6: Monitor Comprehension
Final check on your understanding:
- What do you understand clearly?
- What do you understand partially?
- What are you confused about?
- What don't you know that you don't know? (Can you even guess?)
COMPREHENSION MAP:
- CLEAR: [what you understand well]
- PARTIAL: [what you understand roughly but couldn't explain precisely]
- CONFUSED: [what you're stuck on]
- UNKNOWN UNKNOWNS: [areas where you suspect gaps but can't specify them]
Step 7: Metacognitive Summary
WHAT I WAS DOING: [original strategy]
WHAT I FOUND: [assessment result]
WHAT I'M DOING NOW: [continue / adjust / switch]
BIASES CAUGHT: [list or "none detected"]
COMPREHENSION LEVEL: [high / medium / low / mixed]
NEXT CHECKPOINT: [when to re-run /mtcg]
Integration
Use with:
/prcp -> Improve perception as an alternative cognitive strategy/rskl -> Check reasoning quality when metacognition reveals logical issues/jdgm -> Apply structured judgment when metacognition says "just decide"/sdc -> Deeper self-diagnostic if metacognition reveals persistent patterns/ecal -> Calibrate confidence levels surfaced during bias check
