Menu
Skills in this repository
SkillsMP has collected 1,021 skills from protosphinx/sphinxstack. Open a skill to review its source and details.
protosphinx/sphinxstackShowing 40 of 1,021 collected skills.
Harden a CLI for automation with stable contracts, noninteractive operation, structured output, exit semantics, idempotency, timeouts, security, and compatibility. Use when scripts and agents depend on a command.
Harden dependency updates with provenance, review, compatibility, generated-lock control, staged testing, rollback, and monitoring. Use when automated or manual upgrades can alter production supply-chain risk.
Identify missing values across nulls, blanks, sentinels, absent records, and conditional requirements. Use when assessing dataset completeness before cleaning, modeling, reporting, or migration.
Translate real work, study, care, service, and community experience into portable skill evidence. Use when someone is changing field, returning to work, or cannot see how prior experience fits a target role.
Implement a circuit breaker with meaningful failure classification, bounded state, safe fallback, probing, recovery, and observability. Use when repeated dependency calls can amplify an outage.
Implement a rate limiter with fair identity, atomic accounting, burst policy, distributed consistency, failure behavior, response semantics, and observability. Use when resource use or abuse must be bounded.
Implement a saga workflow with durable state, idempotent steps, explicit compensation, timeout, human intervention, and reconciliation. Use when one business operation spans services without one transaction.
Implement account deletion across identity, product data, derived stores, vendors, retention duties, and user confirmation. Use when users need a reliable way to close an account and remove or irreversibly anonymize eligible data.
Implement an outbox pattern that atomically records domain change and event intent, then publishes with idempotency, ordering, observability, and reconciliation. Use when database state and messages must not diverge.
Implement trustworthy audit logging for consequential actions, security decisions, and administrative changes. Use when an organization must reconstruct who did what, where, when, and with what result.
Implement a browser-storage migration with versioned schemas, privacy controls, idempotent conversion, cross-tab coordination, rollback, and corrupted-state recovery. Use when changing cookies, local storage, session storage, IndexedDB, or cached client data.
Implement client-side caching with explicit ownership, freshness, invalidation, isolation, and recovery. Use when browser data reuse can improve latency or resilience without misleading users.
Implement a staged Content Security Policy from an evidence-based resource inventory, with enforceable directives and safe rollout controls. Use when reducing script, framing, injection, or data-exfiltration risk in a web application.
Implement data lineage from sources through transformations, storage, metrics, models, reports, and consumers with versioned field and run evidence. Use when teams must trace where data came from, how it changed, and what a defect affects.
Implement distributed tracing with interoperable context propagation, useful spans, safe attributes, and measurable sampling. Use when requests cross service or asynchronous boundaries that logs alone cannot reconstruct.
Implement graceful shutdown that stops admission, drains work, preserves durable state, releases resources, and terminates within platform limits. Use when services or workers receive planned or forced termination.
Implement HTTP security headers from an application threat model, route inventory, browser behavior, and staged enforcement evidence. Use when hardening a website or web application against framing, type confusion, leakage, downgrade, or unsafe browser…
Implement idempotency for retried operations using stable request identity, atomic outcome storage, conflict rules, expiry, and recovery tests. Use when duplicate execution could create repeated side effects.
Implement idempotent write requests that survive client retries, timeouts, concurrency, and process failure. Use when repeating an operation must not duplicate its business effect.
Implement standards-based federated login with validated identity claims, protected redirects, and safe account linking. Use when an application must authenticate people through an external identity provider.
Implement an offline-first interface with explicit freshness, queued writes, conflict handling, and recovery. Use when users must read or change data through unreliable connectivity.
Implement optimistic concurrency control with version identity, conditional writes, conflict semantics, retries, merge policy, and user recovery. Use when concurrent updates must not silently overwrite one another.
Implement progressive image loading with responsive sources, reserved layout, meaningful placeholders, priority, lazy loading, accessibility, and failure handling. Use when images delay rendering, waste bandwidth, shift layout, or fail on constrained networks.
Implement responsive images with accurate intrinsic size, source selection, art direction, formats, and accessibility. Use when one image must serve varied layouts, densities, networks, or crops.
Implement secure web sessions across creation, rotation, expiration, revocation, recovery, and distributed enforcement. Use when an application must protect authenticated browser state beyond the initial login.
Implement server-side rendering with deterministic hydration, correct status and metadata, safe caching, and observable failure behavior. Use when routes need useful first responses for users, crawlers, sharing, or performance.
Implement web push with contextual permission, secure subscriptions, user controls, safe delivery, and lifecycle cleanup. Use when timely browser notifications provide clear value after the user leaves a site.
Verify webhook authenticity, freshness, and replay safety using raw request bytes and provider-specific signature rules. Use when accepting externally initiated events that can change application state.
Diagnose an ethical signup, purchase, application, or activation funnel and test one evidence-backed reduction in user friction. Use when people enter a web journey but fail to complete it, conversion rates fall, or a team wants growth without dark patterns…
Shorten and strengthen the path from repository access to a verified first change by repairing setup, documentation, permissions, and feedback loops. Use when new developers stall, rely on oral knowledge, or cannot reproduce the supported environment.
Inspect a web cookie's purpose, scope, attributes, lifetime, consent, transport, and server behavior. Use when debugging sessions, privacy choices, cross-site flows, or unexpected persistence.
Inspect a Content Security Policy report by validating source, directive, blocked resource, document context, browser behavior, privacy, repetition, and exploit relevance. Use when policy violations need triage.
Inspect a CSV file for encoding, delimiter, headers, shape, types, missing values, quoting, duplicates, and sensitive content. Use when receiving or troubleshooting a tabular text export before analysis or import.
Inspect a database query plan through operators, estimates, actual work, memory, I/O, and locking evidence. Use when a query is slow, unstable, or unexpectedly expensive.
Inspect a design token for semantic name, source, aliases, modes, units, consumers, contrast, platform transformation, and deprecation. Use when a visual or behavioral value is inconsistent.
Inspect a metric's meaning, shape, labels, units, aggregation, and collection limits before interpreting a chart. Use when a telemetry value appears surprising or supports an operational decision.
Inspect a process environment without exposing secrets, changing state, or confusing configured and inherited values. Use when runtime behavior may depend on environment, working directory, locale, paths, or limits.
Inspect a request correlation ID across gateway, service, queue, job, and database evidence to reconstruct one transaction. Use when a distributed failure must be traced without broad log searching.
Inspect a web service worker's registration, scope, lifecycle, caches, fetch behavior, update state, and offline effects. Use when a site serves stale content, fails offline, loops updates, or behaves differently after deployment.
Inspect a stack trace to identify the failing path, first relevant application frame, runtime context, and next discriminating check. Use when an error includes a call sequence but the cause is unclear.