All authors
Claude Skills by Amey-Thakur
github.com/Amey-Thakur1,001 skills16 installs1,473 views
- Query Plan ReadingRead EXPLAIN ANALYZE output to find the real cause of a slow query and fix the right thing. Use when a query is slow and you need to know why before changing indexes or SQL.Votes: 0GitHub stars: 7
- Schema DesignDesign database schemas that stay correct under growth and change. Use when creating tables, modeling relationships, or planning schema migrations.Votes: 0GitHub stars: 7
- Sharding PartitioningScale databases by partitioning and sharding with a good key, handling resharding and cross-shard queries, and knowing when to avoid it. Use when one database can no longer hold or serve the load.Votes: 0GitHub stars: 7
- Sql OptimizationDiagnose and fix slow SQL with the query plan as evidence, not folklore. Use when a query is slow, a table scan appears, or database load climbs.Votes: 0GitHub stars: 7
- Time Series DataStore time-series data with retention downsampling, compression, out-of-order handling, and the right store choice. Use when handling metrics, events, or sensor streams that grow relentlessly by time.Votes: 0GitHub stars: 7
- Transactions IsolationChoose isolation levels by the anomalies they prevent, understand locking versus MVCC, and retry on serialization failures. Use when transactional correctness matters or concurrency bugs appear under load.Votes: 0GitHub stars: 7
- Alerting DesignDesign alerts that fire on the symptoms users feel, using error-budget burn rate, a page-worthiness test, and a runbook link on every rule. Use when writing or pruning alerts and you want the pager to mean something instead of crying wolf.Votes: 0GitHub stars: 7
- Binary Search DebuggingHalve the search space repeatedly across code, data, time, and configuration until a single change isolates the failure. Use when the fault could live anywhere across a large surface and reading it all in order is too slow.Votes: 0GitHub stars: 7
- Browser DevtoolsDebug a slow or broken web page with the browser's built-in tools: the network waterfall, the performance panel, and source maps back to original code. Use when a page loads slowly, janks while running, or throws in minified bundle code you cannot read.Votes: 0GitHub stars: 7
- Core DumpsExtract the cause of a crash from a core dump using symbols, backtraces, and post-mortem inspection. Use when a native process died with a signal and left a core file, or when you have a crash dump but no live process to attach to.Votes: 0GitHub stars: 7
- Dashboard DesignBuild dashboards where each panel answers exactly one question, laid out by the RED method for services and the USE method for resources. Use when a board has grown into a wall of graphs nobody can read while an incident burns.Votes: 0GitHub stars: 7
- Deadlock AnalysisFind why threads are stuck forever by dumping their state and building the wait-for graph that reveals the lock cycle. Use when a program hangs with no progress and no crash, or throughput drops to zero while CPU sits idle.Votes: 0GitHub stars: 7
- Debugger FluencyDrive a real debugger with breakpoints, watch expressions, and conditional stops to read live program state at the moment of failure. Use when a bug needs you to see the call stack and variables as it breaks, not reconstruct them afterward.Votes: 0GitHub stars: 7
- DebuggingFind the root cause of a bug with a hypothesis-driven loop instead of guess-and-patch. Use when something fails, crashes, or behaves wrongly and the cause is not yet known.Votes: 0GitHub stars: 7
- Distributed TracingFollow one request across service boundaries using spans, propagated context, and critical-path reading to locate where latency and errors originate. Use when a request is slow or failing and the cause lives between services, not inside any single one.Votes: 0GitHub stars: 7
- Error TrackingWire an error tracker to releases, tune grouping rules, and route each issue to an owner so exceptions become triaged work. Use when exceptions vanish into log noise and nobody knows which are new, which are spiking, or whose they are.Votes: 0GitHub stars: 7
- Flaky Test DiagnosisTurn a test that fails at random into one that fails on demand by controlling the seed, order, and clock. Use when a test passes on re-run, fails only in CI, or blocks a merge for reasons no one can reproduce.Votes: 0GitHub stars: 7
- Git BisectFind the exact commit that introduced a regression by driving git bisect with an automated pass/fail script. Use when a behavior worked in an older build and you need the first bad commit out of hundreds, not a guess.Votes: 0GitHub stars: 7
- HeisenbugsCatch timing-dependent bugs that vanish under observation by amplifying the race and moving logging off the critical path. Use when a fault appears intermittently and disappears the moment you add a print or attach a debugger.Votes: 0GitHub stars: 7
- Log AnalysisMine logs for the cause of a failure by pivoting on correlation ids, clamping the time window, and querying structured fields instead of scrolling. Use when production broke and the logs are the only record of what the system actually did.Votes: 0GitHub stars: 7
- Log LevelsAssign log levels deliberately so error wakes a human, warn flags a trend, and info and debug explain later without burying the signal. Use when logs are either silent during real failures or so noisy that nobody trusts them.Votes: 0GitHub stars: 7
- Memory LeaksFind the object that never gets freed by comparing heap snapshots and following retention paths. Use when a process grows in memory over time, gets OOM-killed, or slows under a garbage collector that keeps working harder.Votes: 0GitHub stars: 7
- Metrics InstrumentationInstrument code with counters, gauges, and histograms picked to answer a specific operational question. Use when you need to watch a system's behavior over time and want the right instrument instead of a wall of unreadable numbers.Votes: 0GitHub stars: 7
- Network DebuggingDiagnose failing or slow network calls by inspecting the actual bytes on the wire with curl, tcpdump, and TLS handshake analysis. Use when a request fails, hangs, or returns the wrong thing and you cannot tell whether the client, the network, or the server is at fault.Votes: 0GitHub stars: 7
- ObservabilityInstrument software so production questions get answered from signals, not guesses. Use when adding logging, metrics, tracing, or alerts, or when a system is hard to debug in production.Votes: 0GitHub stars: 7
- Off By One ErrorsCatch fencepost bugs by writing down the boundary convention and testing the endpoints instead of the middle. Use when a loop, slice, index, or range is off by a single element, drops the last item, or reads one past the end.Votes: 0GitHub stars: 7
- Postmortem DebuggingReconstruct what happened from the artifacts a dead incident left behind: logs, metrics, dumps, and a timeline, when there is no live system left to poke. Use when the outage is over, the process is gone, and all you have is what was written down while it burned.Votes: 0GitHub stars: 7
- Print DebuggingPlace labeled, greppable print statements at decision points to trace the real execution, then remove them cleanly. Use when a debugger is unavailable or awkward and you need to watch values flow through the actual run.Votes: 0GitHub stars: 7
- Production DebuggingDebug a live system without making the incident worse by staying read-only first, gating changes behind flags, and testing on mirrored traffic. Use when a bug only manifests in production and you must investigate against real users and real data.Votes: 0GitHub stars: 7
- Profiling CpuFind the code that actually burns CPU time using a sampling profiler and a flame graph instead of guesswork. Use when a program is slower than it should be and you need to locate the hot path before touching any code.Votes: 0GitHub stars: 7
- Profiling MemoryAttribute memory growth and allocation churn to the exact call sites that produce it using an allocation profiler. Use when a process grows without bound, spends too much time in garbage collection, or allocates far more than its working set explains.Votes: 0GitHub stars: 7
- Race ConditionsDiagnose data races by reasoning about happens-before order, running sanitizers, and stressing the timing until the bug shows. Use when a bug appears only under load, only sometimes, or vanishes when you add a print statement.Votes: 0GitHub stars: 7
- Reproduction FirstBuild a reliable, minimal reproduction before you write a fix so you can prove the bug is actually gone. Use when a report is vague, intermittent, or "works on my machine" and you need solid ground under the debugging.Votes: 0GitHub stars: 7
- Rubber Duck ProtocolExplain the failing code line by line in writing until the sentence you cannot finish exposes the false assumption. Use when you are stuck, re-reading the same code, sure it should work, and it does not.Votes: 0GitHub stars: 7
- Scientific DebuggingDebug by turning a belief about the code into a falsifiable hypothesis, predicting an observable, and running the one probe that can refute it. Use when a bug resists guesswork and shotgun edits are making the code murkier instead of the cause clearer.Votes: 0GitHub stars: 7
- Stack Trace ReadingRead a stack trace to find the cause frame instead of stopping at the symptom on top. Use when an exception, panic, or error dump lands and you need to locate the line that is actually wrong.Votes: 0GitHub stars: 7
- Structured LoggingEmit machine-readable key-value events instead of prose sentences, with a stable schema and controlled field cardinality. Use when logs need to be queried and aggregated, not just read one line at a time by a person.Votes: 0GitHub stars: 7
- Time Travel DebuggingRecord an execution once and replay it deterministically so you can step backward to the moment a value went wrong. Use when a bug is hard to reproduce or the failure surfaces long after its cause, and rerunning changes the outcome.Votes: 0GitHub stars: 7
- Attention MechanismUnderstand what attention computes and why its cost grows quadratically with sequence length, to reason about context limits and efficiency work. Use when working with transformer models or evaluating long-context claims.Votes: 0GitHub stars: 7
- BackpropagationUnderstand how gradients flow backward through a network so vanishing gradients, dead units, and exploding losses become diagnosable. Use when training does not converge or a network learns nothing.Votes: 0GitHub stars: 7
- Batch Size EffectsChoose batch size understanding its effect on gradient noise, memory, throughput, and generalisation, and adjust the learning rate with it. Use when scaling training or running out of memory.Votes: 0GitHub stars: 7
- Learning Rate SchedulesSet and vary the learning rate over training, since it is the hyperparameter that most determines whether a run converges. Use when loss diverges, plateaus early, or oscillates.Votes: 0GitHub stars: 7
- Loss Function SelectionChoose a loss that matches the task and the metric you care about, and understand what each penalises. Use when a model optimises well and performs badly on the thing that matters.Votes: 0GitHub stars: 7
- Optimizer SelectionChoose an optimiser and its hyperparameters based on the problem rather than habit, and know what each actually does. Use when training is unstable, slow to converge, or generalising poorly.Votes: 0GitHub stars: 7
- Overfitting DiagnosisDistinguish overfitting from underfitting, data leakage, and distribution shift, since all four present as poor validation performance. Use when validation performance is worse than expected.Votes: 0GitHub stars: 7
- Regularization TechniquesReduce overfitting with weight decay, dropout, augmentation, and early stopping, choosing by why the model is overfitting. Use when validation performance diverges from training performance.Votes: 0GitHub stars: 7
- Scaling LawsReason about how model performance improves with parameters, data, and compute, to allocate a training budget sensibly. Use when planning a training run or evaluating claims about model size.Votes: 0GitHub stars: 7
- TokenizationUnderstand how text becomes tokens, and why token boundaries explain model behaviour on numbers, code, and non-English text. Use when a model behaves strangely on specific strings or costs more than expected.Votes: 0GitHub stars: 7
- Training Loop DesignStructure a training loop with correct ordering, evaluation, checkpointing, and logging so runs are debuggable and resumable. Use when writing or reviewing training code.Votes: 0GitHub stars: 7
- Weight InitializationInitialise parameters so signal and gradients propagate at usable scale from the first step. Use when a deep network fails to train from the start or diverges immediately.Votes: 0GitHub stars: 7