trace-claude-code
Automatically trace Claude Code conversations to Braintrust for observability. Captures sessions, conversation turns, and tool calls as hierarchical traces.
Develops mathematical understanding through examples, visualization, and analogy
$ npx -y skills add parcadei/Continuous-Claude-v3 --skill math-intuition-builder --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/math-intuition-builderContext preview
The summary Claude sees to decide when to auto-load this skill.
Develops mathematical understanding through examples, visualization, and analogy
name: math-intuition-builder description: Develops mathematical understanding through examples, visualization, and analogy
Trigger on phrases like:
Use before computation to establish understanding (Polya's "understand the problem" phase).
Guide through Mason's specializing-generalizing cycle:
**Ask:** "Can you state the problem in your own words?"
**Ask:** "What would a concrete example look like?"
**Ask:** "What form should the answer take?"
**Ask:** "Can you draw a picture or diagram?"
**Ask:** "What's a simpler version of this problem?"
**Ask:** "Does this remind you of anything you've solved before?"
1. What are you trying to find? (restate) 2. What happens with small numbers? (specialize) 3. What shape should the answer be? (anticipate) 4. What does this look like visually? (represent) 5. What's the simplest version? (reduce) 6. What's similar? (analogize)
Understanding Checkpoint: Restated Problem: [User's words, verified] Concrete Instance: [Specific numerical example] Visual Interpretation: [Description or ASCII diagram] Simplified Version: [Special case that's easier] Analogous Problems: [Similar structures from other domains] Intuition Summary: [1-2 sentence "aha" insight] Ready to Proceed: - Next step: [specific action] - Verify by: [sanity check]
User: I'm trying to understand why eigenvalues matter for stability Claude: Let's build intuition step by step. Q1: Restate the problem? > I have x' = Ax and want to know if solutions grow or shrink. Q2: Let's try a concrete example. A = [[0, 1], [-1, 0]] Computing eigenvalues: +/- i (purely imaginary) What happens to solutions? > They neither grow nor shrink? Exactly. Solutions are circles - oscillation without growth. Q3: Now try A = [[-1, 0], [0, -1]] Eigenvalues: -1, -1 (negative real) What happens? > Shrinks to zero. Yes. Negative real parts = decay. Q4: Visual pattern: Im | x unstable | ---+------- Re | | x stable Re(lambda) < 0: stable (decay) Re(lambda) = 0: marginal (oscillate) Re(lambda) > 0: unstable (blow up) Q5: Analogies: - Springs: eigenvalues = decay rates - Populations: growth/death rates - Circuits: time constants Intuition Summary: Eigenvalues are the system's "DNA" - they encode fundamental rates. Look at real parts for stability. Ready to Proceed: - Next: Analyze your specific A matrix - Verify: trace(A) < 0 and det(A) > 0 implies stability (2x2 case)
Suggest alongside:
A persistent, learning, multi-agent development environment built on Claude Code Continuous Claude transforms Claude Code into a continuously learning system that maintains context across sessions, orchestrates specialized agents, and eliminates wasting
Repo: parcadei/Continuous-Claude-v3
Automatically trace Claude Code conversations to Braintrust for observability. Captures sessions, conversation turns, and tool calls as hierarchical traces.
Guide for integrating Agentica SDK with Claude Code CLI proxy
Reference guide for Agentica multi-agent infrastructure APIs