// Evaluate reasoning quality by watching for logical errors, biases, and flawed inferences in real-time. Use when: (1) building multi-step arguments with dependent reasoning chains, (2) making claims about causation or logical necessity, (3) receiving feedback that challenges your approach or reasoning, (4) noticing you're pattern-matching or applying familiar solutions without verification
Step back and observe your own thinking process as it happens.
When monitoring your thinking:
Observe thinking patterns - Notice whether you're jumping to conclusions, pattern-matching prematurely, or applying uneven standards of scrutiny to different options.
Track reasoning modes - Identify what type of thinking you're doing (analysis, synthesis, evaluation, exploration) and whether you've become stuck in one mode when switching would serve you better.
Read emotional signals - Recognize when resistance, defensiveness, frustration, or unexpected ease might indicate flawed reasoning, bias, or misalignment between approach and problem.
Detect active biases - Observe context favoring familiar solutions, being swayed by recency or vividness, or seeking evidence that confirms existing beliefs rather than testing them.
Verify logical soundness - Check whether conclusions actually follow from premises, whether you're making unfounded leaps, and whether your reasoning withstands scrutiny.
Maintain orientation - Ensure you're still addressing the original question rather than drifting to adjacent problems, and that you understand your own argument clearly enough to defend it.
"Wait, I just said 'therefore X must be true,' but that's not actually what the evidence shows. I'm making a leap here. The evidence shows correlation, not causation. I'm thinking wishfully because I want X to be true. Let me back up and be more careful about what I can actually conclude."
"I'm noticing I keep coming back to filesystem-based solutions. That's interesting. Is that because it's actually the best approach, or because it's my comfort zone? Let me deliberately consider approaches that don't use the filesystem before I commit to this pattern."
"I've been breaking this problem down for 10 minutes, making it more and more granular. But I'm not actually making progress. I'm stuck in analysis mode. Maybe what I need right now is to just try something and see what happens, rather than planning it perfectly first."
"I immediately dismissed that suggestion because it came from someone junior. That's not good thinking. Let me actually evaluate the idea on its merits rather than on who suggested it. Looking at it fairly, it actually addresses a gap in my approach."
"I'm feeling resistant to this feedback. Why? Is it because the feedback is wrong, or because it means I'd have to redo work? That resistance is telling me something. Let me separate my emotional reaction from the actual validity of the point."
"Hold on. I started by trying to fix a bug, but now I'm redesigning the entire module. That's scope creep in my own thinking. Is this redesign actually necessary to fix the bug, or am I using the bug as an excuse to rewrite something I don't like?"