Connect your Claude / AI to 2,500+ algorithms, data structures and systems

tirogla is an MCP server that your AI calls while it designs and builds. Name a system and you get its component graph: its parts and what each part is built on, down to the basic algorithms, with alternatives. Every page opens with a one-pager (the problem, how it works) and goes on to the deep dive: code, complexity, pitfalls, efficiency and the libraries that implement it.

How it works

  1. Connect. claude mcp add --transport http tirogla https://tirogla.com/mcp
  2. Say what you want to build. "A news harvester for every RSS feed", "a chess engine", "an inference server for an open model". Your AI finds the matching systems and the parts each one needs.
  3. Let your coding agent execute our specs. It builds the system in small tested steps, citing the pages it used.

What your AI gets when you want to build a transformer-based chess engine

Next to the system: the parts it is made of, like Monte Carlo tree search. Each column further right: what the parts before it are built on, like embeddings. Click a box to open its page.

Open the neural-chess-engine pageneural-chess-engine Batched MCTS inferencegame-ai: Feed an accelerator from tree search driven by a learned evaluator: virtual loss gathers leaves within a tree, parallel searches fill batches, an async queue overlaps CPU and GPU.›Legal move generationgame-ai: Lists every legal move from bitboard attack tables, drops those that leave the own king in check, and is proven correct by perft node counts.›Monte Carlo tree searchgame-ai: Grows a search tree one simulation at a time, picking visits by a bandit rule (UCT or PUCT) and backing up averages; the search behind AlphaZero.›AlphaZero self-playmachine-learning: A policy-value net guides PUCT search; search visit counts and game results become its training targets, so it learns from its own improved play.›Transformer policy-valuenetworkmachine-learning: One encoder-only transformer reads the board as 64 square tokens and outputs move probabilities and a win/draw/loss estimate for a tree search.›Neural network trainingml-training: The end-to-end training loop, from batch, forward, loss, backward, clip, AdamW with warmup-cosine schedule, mixed precision, checkpoints, evaluation and data-parallel scaling.›MPMC bounded queueconcurrency: Vyukov's array queue for many producers and consumers: each slot carries a sequence number, so one CAS per operation claims a slot without locks.›Dynamic batchingml-training: Queue independent requests and dispatch them as one batch when it is full or the oldest has waited max_delay; bounded extra latency buys throughput.›Bitboardsgame-ai: One 64-bit word per set of up to 64 items, so set algebra becomes shifts, ANDs and table lookups; magic bitboards give ray queries in one multiply.›Zobrist hashinggame-ai: A state's hash is the XOR of one random 64-bit key per (item, slot) fact that holds in it, updated in a few XORs per change.›Transposition tablegame-ai: A fixed-size, lossy hash table of search results keyed by a state hash; gives ordering and cutoffs for repeated states, shared lock-free between threads.›Multi-armed banditsmachine-learning: Choose among K options repeatedly while learning their payoffs: epsilon-greedy, UCB1 and Thompson sampling keep total regret low where a fixed split or a fixed exploration rate pays forever.›PRNGs (PCG, xoshiro)numerics: Small, fast, statistically strong non-cryptographic generators: PCG32 and xoshiro256**, with seeding, parallel streams and unbiased ranges.›Experience replaymachine-learning: A fixed-size ring buffer of past experience, sampled uniformly or by priority, so an online learner trains on decorrelated mini-batches and reuses each item.›Symmetry-group augmentationmachine-learning: Transform each input and its target together with a symmetry group to multiply training data, or average predictions over the group at test time.›Attention variantsmachine-learning: Multi-head, multi-query, grouped-query, sliding-window and latent attention; same softmax(QKᵀ/√d)V core, different trade-offs between quality and KV-cache size.›Embeddingsmachine-learning: A learned table mapping discrete ids to dense vectors: gather forward, sparse scatter-add backward, row-wise optimizers to keep huge tables cheap.›Rotary position embedding(RoPE)machine-learning: Encode position by rotating pairs of query and key dimensions by position-dependent angles, so attention scores depend only on relative distance; the LLM default.›+ 9 moreFalse sharingconcurrency: Independent variables written by different cores on one cache line force the line to bounce between cores; pad or regroup hot data to fix it.›Backpressuresystems: Bounded queues and credit-based flow so slow consumers throttle producersHierarchical timing wheelsystems: Timers bucketed by expiry into wheels of slots, coarser wheels for far deadlines: O(1) insert and cancel for millions of timeouts.›Memory orderingconcurrency: What Relaxed, Acquire, Release and SeqCst atomics actually guarantee, what they cost on x86 and ARM, and which one each pattern needs.›Bayesian rate estimationstatistics: Estimate a conversion or event rate from few observations with a Beta-Binomial or Gamma-Poisson update: two counters, an honest interval, and P(B beats A) directly.›Reverse-mode automaticdifferentiationml-training: Record the forward computation on a tape, then sweep it backwards once to get the gradient of a scalar loss for all inputs.›+ 9 moreCPU caches and the memoryhierarchyhardware-os: Sizes, line widths and latencies of L1/L2/L3/DRAM on modern x86 and Apple M-series, and the data-layout rules that follow from them.›Slab allocatorsystems: Pre-carve pages into equal-size slots per size class with an in-place free list: O(1) alloc and free, no external fragmentation, cache-warm reuse.›+ 4 more
See all systems →

Get Pro access

Free
€0forever

Search and read the library.

Start free

Included:

  • Search across all systems and pages
  • Overviews, How it works, Code, Complexity, Variants, Sources
  • Documentation requests
  • About 100 calls per day
Pro monthly
€20/ month

Billed monthly. Cancel any time.

Get Pro monthly

Everything in Free, plus:

  • System blueprints: every part a system needs
  • Design questions with recommended answers
  • Efficiency, Implementation notes, Pitfalls, In production
  • Library verdicts, live health, security advisories
  • Build bundle for your AI
  • Higher limits and commercial use
Early access −50 %
Pro yearly
€16€8/ month

€96 for the first year instead of €192, then €192 a year. Early-access deal until 15 Nov.

Get Pro yearly

All Pro features, plus:

  • One payment a year
  • 60 % less than paying monthly in the first year