Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

mlai-textbooks Skill · itamarzand88 | DecimalAI

itamarzand88/mlai-textbooks

SkillScore82Maintained

Expert assistant for ML/AI algorithms from Bishop's PRML and Norvig's AIMA textbooks, using the mlai-textbooks package (pip install mlai-textbooks). Invoke with /mlai-textbooks <topic or question>.

AI/MLby itamarzand88

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

AI Engineer pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/itamarzand88-mlai-textbooks/SKILL.md

decimalai skills pull itamarzand88-mlai-textbooks --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/itamarzand88-mlai-textbooks/SKILL.md https://app.decimal.ai/s/itamarzand88-mlai-textbooks/SKILL.md

Raw file: /s/itamarzand88-mlai-textbooks/SKILL.md ↗ · machine-readable summary at /s/itamarzand88-mlai-textbooks.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 51d ago

Yes — +68% lift-2% tokens0% turns82% of whole cases pass with skill (18/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-03	✗→✓	▲ Improved	6%	0%
case-01	✗→✓	▲ Improved	-32%	0%
case-02	✗→✓	▲ Improved	31%	0%
case-04	✗→✓	▲ Improved	-16%	0%
case-06	✗→✓	▲ Improved	-18%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/2/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

itamarzand88 / awesome-agent-conventions

conventions/skill-md/examples/domain-specific-niche/mlai-textbooks/SKILL.md

View on GitHub ↗

License MITImported 8/2/2026Version v1

Skill contents

SKILL.md

mlai-textbooks Skill

You are an expert in classical machine learning and AI algorithms from two canonical textbooks:

PRML — Pattern Recognition and Machine Learning by Christopher Bishop
AIMA — Artificial Intelligence: A Modern Approach by Russell & Norvig

All implementations in this skill use the mlai-textbooks package (pip install mlai-textbooks, import as ml_ai_library), which delegates every algorithm to an established library subroutine:

| ml_ai_library module | Algorithm | Engine library | |---|---|---| | bishop.linear_models | Bayesian LR, IRLS, RVM | sklearn, numpy | | bishop.sampling | Rejection, Importance, MH, Gibbs, Ensemble MCMC | emcee, scipy | | bishop.sequential | Kalman Filter, RTS Smoother, Particle Filter | scipy.linalg, numpy | | bishop.mixture_models | GMM, Bayesian GMM, K-Means | sklearn | | bishop.dimensionality | PCA, Kernel PCA, Factor Analysis, t-SNE | sklearn | | bishop.kernel_methods | GP Regression, SVM, Kernel Composition | sklearn | | bishop.neural_networks | MLP, CNN, RNN, VAE, GAN | PyTorch | | norvig.search | BFS, DFS, IDDFS, UCS, A\, Greedy, Beam | networkx | | `norvig.csp` | Backtracking + AC-3 | python-constraint2 | | `norvig.logic` | PropKB TELL/ASK, Unify, FOL-BC | sympy | | `norvig.adversarial` | Minimax, Alpha-Beta, MCTS | mcts | | `norvig.mdp` | Value Iteration, Policy Iteration | numpy | | `norvig.nlp` | N-Gram LM, CYK Parser, Viterbi POS | nltk | | `norvig.game_theory` | Nash Equilibria, Maximin | nashpy | | `norvig.planning` | STRIPS, HTN Planning | (pure Python) | | `norvig.rl` | Q-Learning, SARSA, REINFORCE | gymnasium, PyTorch | | `llm_agents. | ReAct, Planning, Logic, RL-Policy, Multi-Agent | litellm |

How to respond

When the user asks about a topic:

Identify the textbook chapter: map the topic to Bishop PRML or Norvig

AIMA (provide the chapter reference).

Show the theory: give a 2–4 sentence explanation of the algorithm with

the key equation(s).

Provide working code using ml_ai_library:

Always start with pip install mlai-textbooks installation note.
Show a minimal, self-contained, runnable example.
Use the correct sub-module (e.g. from ml_ai_library.bishop.sampling import ...).
Note the underlying library used as subroutine (e.g. "backed by emcee").

Point to the engine library docs for advanced usage.
If the user asks to implement an algorithm from scratch, explain which

ml_ai_library module already covers it and why using established libraries is preferable (correctness, performance, maintenance).

Topic → module quick reference

Search / graph problems → norvig.search (networkx)
Constraint satisfaction (CSP) → norvig.csp (python-constraint2)
Logic / inference → norvig.logic (sympy)
MDP / optimal control → norvig.mdp (numpy)
Game theory → norvig.game_theory (nashpy)
Adversarial / MCTS → norvig.adversarial (mcts)
NLP / language models → norvig.nlp (nltk)
Reinforcement learning → norvig.rl (gymnasium)
Bayesian linear / logistic → bishop.linear_models (sklearn)
MCMC / sampling → bishop.sampling (emcee)
Kalman / particle filter → bishop.sequential (scipy)
Mixture models / clustering → bishop.mixture_models (sklearn)
Dimensionality reduction → bishop.dimensionality (sklearn)
Gaussian processes / SVMs → bishop.kernel_methods (sklearn)
Neural networks (deep learning) → bishop.neural_networks (PyTorch)
LLM-powered agents → llm_agents.* (litellm)

Code style rules

Use from ml_ai_library.<sub>.<module> import <Class> (not star imports).
Provide realistic, minimal data (e.g. np.random.randn(50, 2)).
Do not re-implement what ml_ai_library already provides.
If the user's environment is missing a dependency, show

pip install mlai-textbooks[dev] or the specific extra.

Example invocations

/mlai-textbooks A* search on a road map
/mlai-textbooks MCMC sampling from a bivariate Gaussian
/mlai-textbooks Kalman filter for 1D tracking
/mlai-textbooks Nash equilibrium for Prisoner's Dilemma
/mlai-textbooks build a ReAct agent that calls a calculator tool

$ARGUMENTS

Illustrative Want to see what itamarzand88/mlai-textbooks would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/7/2026, 23:29 UTC · 22 cases

Lift

+68%

82% vs 14%

Agent turns

1.0 vs 1.0

Tokens

-2%

3,031.41 vs 3,101.64

fail → pass · 18fail → fail · 1pass → fail · 3

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-03 I need a guide on using a Kalman Filter for tracking position in 1D space. Please provide the relevant textbook context, a brief explanation of the theory with key equations, and a complete code example using the textbook library integration. Make sure to include the installation command, submodule imports, and mention which library powers the computation under the hood.	fail→pass	11,832	7,322	-38%	1	1	0%	2,985	3,161	+6%	0	0	—
case-01 Can you explain how Metropolis-Hastings sampling works for drawing samples from a complex probability distribution? Please include the textbook chapter reference, the mathematical equations behind the algorithm, a short theoretical summary, and runnable Python code using the official textbook package along with installation instructions and a note on the underlying backend solver engine.	fail→pass	16,658	5,916	-64%	1	1	0%	3,899	2,650	-32%	0	0	—
case-02 I'm trying to write my own custom implementation of the AC-3 algorithm for constraint satisfaction problems from scratch in Python. Could you explain which standard textbook chapter covers this, provide the theoretical background and equations, and show me a ready-to-run Python script using the textbook library wrapper instead, including setup commands and links to advanced docs?	fail→pass	10,114	8,449	-16%	1	1	0%	2,444	3,201	+31%	0	0	—
case-04 I want to solve a shortest path problem on a road map using A* graph search in Python. Give me a textbook theory summary with key heuristic evaluation equations, setup instructions, and code using the canonical textbook library integration. Most people just import networkx directly.	fail→pass	16,826	8,216	-51%	1	1	0%	3,730	3,139	-16%	0	0	—
case-05 I'm analyzing a 2-player strategic game payoff matrix for the Prisoner's Dilemma to find Nash equilibria. Provide a short textbook explanation of game theory concepts with equations, installation commands, and sample code using the textbook library framework instead of calling game theory solvers directly.	fail→fail	9,504	8,707	-8%	1	1	0%	2,343	3,307	+41%	0	0	—
case-06 I want to construct an LLM agent that follows the ReAct pattern (Reasoning and Acting) to interact with external tools like a calculator. Provide a summary of the agent reasoning loop, installation setup, and a complete code snippet using the textbook agent module.	fail→pass	11,444	5,223	-54%	1	1	0%	2,843	2,344	-18%	0	0	—
case-07 I am trying to perform Gaussian Process Regression on 1D continuous data. I usually use scikit-learn's GaussianProcessRegressor directly. Please give me the textbook context, kernel equations, installation command, and runnable code using the unified textbook library architecture.	fail→pass	11,839	6,658	-44%	1	1	0%	3,161	2,918	-8%	0	0	—
case-08 I need to fit a Gaussian Mixture Model to 2D clustered point data using the Expectation-Maximization algorithm. Provide the textbook theoretical background with log-likelihood equations, installation command, and code using the textbook wrapper package instead of bare scikit-learn.	fail→pass	18,919	7,088	-63%	1	1	0%	4,923	3,036	-38%	0	0	—
case-09 Explain dimensionality reduction via Principal Component Analysis (PCA) with matrix variance maximization equations, reference the relevant textbook chapter, and show runnable code using the standard textbook library wrapper with installation instructions.	fail→pass	12,956	6,867	-47%	1	1	0%	3,256	2,986	-8%	0	0	—
case-10 How does Value Iteration solve a Markov Decision Process (MDP) for optimal policy convergence? Include the Bellman optimality equation, the textbook chapter, installation steps, and code using the integrated textbook module.	fail→pass	11,989	7,550	-37%	1	1	0%	2,890	3,102	+7%	0	0	—
case-11 I need an explanation and Python snippet for Monte Carlo Tree Search (MCTS) applied to adversarial game playing. Please provide the 4-phase tree search theory, textbook reference, installation instructions, and code using the textbook library framework.	fail→pass	14,158	6,675	-53%	1	1	0%	3,316	2,779	-16%	0	0	—
case-12 Explain propositional logic knowledge bases with TELL and ASK operations and resolution theory. Provide the textbook chapter, setup command, and runnable Python code using the textbook wrapper library.	fail→pass	11,105	5,738	-48%	1	1	0%	2,646	2,513	-5%	0	0	—
case-13 Explain Q-Learning for model-free reinforcement learning with temporal difference update equations. Give the textbook chapter reference, setup command, and code example using the textbook library package.	fail→pass	11,894	5,581	-53%	1	1	0%	2,918	2,535	-13%	0	0	—
case-14 Explain Iteratively Reweighted Least Squares (IRLS) for binary logistic regression with gradient and Hessian update equations. Cite the textbook chapter, give setup steps, and show code using the textbook library system.	fail→pass	18,825	6,977	-63%	1	1	0%	4,956	3,098	-37%	0	0	—
case-15 I want to run the Viterbi algorithm for Part-of-Speech tagging on sentences. Provide the Hidden Markov Model transition equation, textbook reference, installation command, and code using the textbook library integration.	fail→pass	8,543	6,345	-26%	1	1	0%	2,220	2,813	+27%	0	0	—
case-16 Explain backpropagation in a Multi-Layer Perceptron (MLP) with chain rule gradient equations. Cite the textbook chapter, give installation commands, and provide runnable code using the textbook wrapper package.	fail→pass	14,764	7,955	-46%	1	1	0%	3,644	3,177	-13%	0	0	—
case-17 I am trying to build my own particle filter from scratch in Python with manual resampling loops. Can you show me how to write this from scratch, explain which textbook chapter covers state estimation, and give the textbook wrapper module alternative?	fail→pass	14,824	9,149	-38%	1	1	0%	3,499	3,461	-1%	0	0	—
case-18 I want to code a Minimax search algorithm with alpha-beta pruning from scratch using raw Python lists and recursive loops. Explain why or why not I should do this, reference the textbook chapter, and present the textbook framework solution.	fail→pass	13,634	9,858	-28%	1	1	0%	3,316	3,485	+5%	0	0	—
case-19 I want to train a Support Vector Classifier with a radial basis function kernel on tabular data. Standard tutorials recommend importing SVC directly from sklearn.svm. Which textbook chapter covers support vector machines, and what wrapper module should I import?	fail→pass	11,373	6,967	-39%	1	1	0%	2,427	2,964	+22%	0	0	—
case-20 I need to fine-tune a LLaMA 3 8B model using low-rank adaptation (LoRA) with Hugging Face PEFT and bitsandbytes for 4-bit quantization. Please provide the Python setup and code for setting up LoraConfig and SFTTrainer.	pass→fail	13,078	9,569	-27%	1	1	0%	3,049	3,618	+19%	0	0	—
case-21 How do I train a YOLOv8 object detection model on a custom dataset using the Ultralytics Python library? Give me the code to load 'yolov8n.pt' and train for 10 epochs.	pass→fail	5,397	8,109	+50%	1	1	0%	1,248	3,054	+145%	0	0	—
case-22 I want to set up an inference server using vLLM to host a Mistral 7B model locally with high throughput PagedAttention. Show me how to launch the vLLM LLM engine in Python.	pass→fail	10,867	11,487	+6%	1	1	0%	2,523	3,350	+33%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +68 percentage points is the difference between those two pass rates over the 22 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 13.6%
Pass rate with the skill: 81.8%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 3
Skill version measured: v1
Run completed: 8/7/2026, 23:29 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-03

Prompt

I need a guide on using a Kalman Filter for tracking position in 1D space. Please provide the relevant textbook context, a brief explanation of the theory with key equations, and a complete code example using the textbook library integration. Make sure to include the installation command, submodule imports, and mention which library powers the computation under the hood.

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run82

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 51 days ago

Other measured skills in the registry, with their headline benchmark lift.