derive
Perform step-by-step mathematical derivations with full explanations and LaTeX output.
Skill by rafeeqinea
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/rafeeqinea-derive
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: documents
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/derive
What's inside
- Skill name
derive- Description
- Perform step-by-step mathematical derivations with full explanations and LaTeX output. Trigger phrases: /derive, "derive this", "show the derivation", "prove that", "mathematical proof", "step by step math"
- License
- MIT
- Instructions
- 183 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- allowed-tools: must be a space-separated string
Files (1)
- SKILL.md
SKILL.md (first lines)
# Mathematical Derivation Skill
## Purpose
Produce rigorous, step-by-step mathematical derivations from premises to conclusion. Every step is justified, every variable is defined, and the output is formatted for direct inclusion in academic reports.
## Supported Domains
- **Linear algebra**: Matrix operations, eigenvalue problems, decompositions, trace/determinant identities
- **Calculus**: Differentiation, integration, Taylor/Maclaurin series, limits, multivariable calculus
- **Probability and statistics**: Bayes' theorem, expectations, distributions, maximum likelihood, Bayesian inference
- **Optimisation**: Gradient descent convergence, Lagrange multipliers, KKT conditions, convexity proofs
- **Differential equations**: ODEs, PDEs, separation of variables, Green's functions, weak formulations
- **Numerical methods**: Discretisation error analysis, stability proofs, convergence rates
- **Machine learning**: Backpropagation, loss function gradients, kernel tricks, information-theoretic bounds
## Workflow
### Step 1: Clarify the Derivation
Before starting, establish:
1. **Starting point**: What are the premises, axioms, or given equations?
2. **Target**: What expression, identity, or result needs to be derived?
3. **Level of rigour**: Sketch (key steps only), standard (all steps, informal justification), or formal (proof-level, every inference rule named)
4. **Notation conventions**: Does the user follow specific notation? (e.g., bold for vectors, hat for estimates, bar for means)
5. **Context**: Is this for a report, homework, or personal understanding? This affects the level of explanation.
### Step 2: Define All Variables
Before the first equation, produce a notation table:
```
Let:
- A ∈ ℝ^{m×n} : the coefficient matrix
- x ∈ ℝ^n : the solution vector
- b ∈ ℝ^m : the right-hand side vector
- ‖·‖_2 : the Euclidean (L2) norm
- κ(A) : the condition number of A, defined as ‖A‖‖A⁻¹‖
```
Every variable used in the derivation must appear in this table. No symbol should appear without definition.
### Step 3: Perform the Derivation
Structure each step as:
```
Step k: [Brief description of what this step does]
[Equation from previous step]
= [New expression] (by [justification])
= [Next expression] (by [justification])
[1-3 sentences explaining WHY this step was taken and WHAT it accomplishes]
```
#### Justification Types (always specify one):
| Justification | When to Use |
|---------------|-------------|
| by definition | Substituting a definition |
| by [Named Theorem] | Applying a known theorem (state it) |
| by [algebraic operation] | Factoring, expanding, combining like terms |
| by substitution of (k) | Substituting result from step k |
| by linearity of [operator] | Using linearity of expectation, integral, etc. |
| by chain rule | Composite function differentiation |
| by product rule | Product differentiation |
| by integration by parts | The standard IBP formula |
| by Cauchy-Schwarz | Applying the inequality |
| by triangle inequality | Applying |a+b| ≤ |a|+|b| |
| by assumption | Something stated in the premises |
| by symmetry of A | Using A = A^T |
| by positive definiteness | Using x^T A x > 0 for x ≠ 0 |
| rearranging | Simple algebraic manipulation |
### Step 4: Handle Branches and Cases
When a derivation requires case analysis:
```
We consider two cases:
Case 1: [Condition]
[Derivation for this case]
∴ [Result for case 1]
Case 2: [Condition]
[Derivation for this case]
∴ [Result for case 2]
Combining both cases: [Final result]
```
### Step 5: Verify the Result
After completing the derivation, perform at least one verification:
- **Dimensional analysis**: Do the units/dimensions match on both sides?
- **Special case check**: Substitute a simple known case (e.g., n=1, A=I) and verify
- **Boundary check**: Does the result behave correctly at extremes (x→0, x→∞)?
- **Numerical check**: Plug in specific numbers and verify both sides agree
- **Symmetry check**: If the result should be symmetric in certain variables, verify it is
State which verification was performed and show the check.
### Step 6: Output Format
Provide the derivation in two formats:
#### A. Annotated Markdown (for understanding)
Each step with full English explanation, motivation, and commentary. Use this for teaching and personal notes.
#### B. LaTeX (for reports)
Clean, publication-ready LaTeX using proper environments:
```latexFrom rafeeqinea/claude-academic-toolkit, MIT.
More from rafeeqinea/claude-academic-toolkit
- 0.409score
explain-file
rafeeqinea
Read a source file and explain every line in depth.
SkillDFreshness: DClaude Code - 0.409score
self-improve
rafeeqinea
Deep self-analysis and improvement.
SkillDFreshness: DClaude Code - 0.409score
reflect
rafeeqinea
End-of-session reflection.
SkillDFreshness: DClaude Code +9 - 0.406score
benchmark
rafeeqinea
Systematic benchmarking and comparison of algorithms, models, or approaches under identical conditions.
SkillDFreshness: DClaude Code +9 - 0.406score
drive-research
rafeeqinea
Search Google Drive for papers, datasets, documents, and pull files into the conversation.
SkillDFreshness: DClaude Code +9 - 0.406score
notebook-query
rafeeqinea
Query NotebookLM notebooks directly from Claude Code for grounded, citation-backed answers.
SkillDFreshness: DClaude Code +9 - 0.406score
notion-notes
rafeeqinea
Push research notes, experiment results, and session summaries to Notion.
SkillDFreshness: DClaude Code +9 - 0.406score
docs-lookup
rafeeqinea
Fetch the latest documentation for any programming library using Context7.
SkillDFreshness: DClaude Code +9 - 0.406score
experiment
rafeeqinea
Run, track, and compare ML experiments with structured logging.
SkillDFreshness: DClaude Code +9 - 0.406score
debug
rafeeqinea
Structured debugging workflow that teaches the debugging process, not just fixes the bug.
SkillDFreshness: DClaude Code +9 - 0.403score
teach
rafeeqinea
Explain any concept from absolute scratch using first principles, analogies, examples, and progressive complexity.
SkillDFreshness: DClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Mar 9, 2026 (7 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from rafeeqinea
- 0.409score
explain-file
rafeeqinea
Read a source file and explain every line in depth.
SkillDFreshness: DClaude Code - 0.409score
self-improve
rafeeqinea
Deep self-analysis and improvement.
SkillDFreshness: DClaude Code - 0.409score
reflect
rafeeqinea
End-of-session reflection.
SkillDFreshness: DClaude Code +9 - 0.406score
benchmark
rafeeqinea
Systematic benchmarking and comparison of algorithms, models, or approaches under identical conditions.
SkillDFreshness: DClaude Code +9 - 0.406score
drive-research
rafeeqinea
Search Google Drive for papers, datasets, documents, and pull files into the conversation.
SkillDFreshness: DClaude Code +9 - 0.406score
notebook-query
rafeeqinea
Query NotebookLM notebooks directly from Claude Code for grounded, citation-backed answers.
SkillDFreshness: DClaude Code +9
Cite this page
derive skill. Agent Mart. https://agentmart.dev/skills/rafeeqinea-derive (last updated 2026-10-24)