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.
Tailor a resume for a leadership role using verified scope, decisions, team and organizational influence, outcomes, governance, and role requirements. Use when applying for manager, director, executive, or functional-lead positions.
Adapt the resume to one specific posting. Vocabulary, ordering, and cuts, nothing invented. Use when they say "tailor my resume", "adjust my resume for this job", or paste a posting they want to apply to.
Capture lecture claims, examples, definitions, questions, and source links in notes designed for later retrieval and correction. Use when attending or reviewing a class, talk, demonstration, or recorded lecture.
Test a cookie banner and preference center for choice, accessibility, storage, tag behavior, withdrawal, locale, and policy consistency. Use when validating consent behavior before or after website changes.
Verify a data migration against source-to-target mappings, invariants, edge cases, reconciliation, and recovery behavior. Use when testing migration code or a rehearsal before changing production data.
Test a data-pipeline backfill for scope, snapshot consistency, idempotency, ordering, capacity, correctness, downstream effects, and rollback. Use when preparing to recompute historical partitions or events in a production data system.
Test a deep link across cold and warm launch, authentication, authorization, routing, fallback, app and web handoff, parameters, and recovery. Use when a URL or app link should open a specific destination safely.
Test a web file download for authorization, response headers, filename, integrity, type, size, content, accessibility, and failure behavior. Use when a user retrieves generated or stored files.
Test a multiregion web failover across routing, compute, state, identity, data, queues, storage, dependencies, observability, recovery, and reconciliation. Use when regional resilience claims need controlled evidence.
Test a URL redirect chain for hop count, status, destination, query preservation, security, caching, and loops. Use when a page, domain, campaign link, or protocol migration does not land correctly.
Test a web form across input, validation, accessibility, submission, failure, duplication, and recovery. Use when a form must reliably create a consequential record or request.
Test a website on representative mobile devices, browsers, networks, orientations, and input conditions. Use when a critical journey must work beyond desktop emulation.
Test a zero-downtime deploy across mixed versions, readiness, drain, schema compatibility, sessions, jobs, traffic, rollback, and user outcomes. Use when a release promises uninterrupted service.
Test an API integration across contract, authentication, state, retries, errors, limits, observability, and cleanup. Use when verifying a new or changed consumer-provider connection before release.
Test an ecommerce checkout across cart, identity, address, delivery, tax, discounts, payment, confirmation, accessibility, and financial reconciliation. Use when releasing or diagnosing a customer purchase journey.
Test authentication journeys across success, failure, abuse, recovery, session, and accessibility conditions. Use when sign-up, login, MFA, logout, or account recovery must be release-ready.
Test critical web journeys across representative browser engines, versions, devices, and capability boundaries. Use when a release must work beyond the team's default browser.
Test a web journey using keyboard-only navigation for reachability, order, focus, activation, escape, and recovery. Use when checking interactive accessibility without a pointer.
Test service shutdown behavior across traffic drain, readiness, in-flight work, queues, transactions, leases, deadlines, and restart recovery. Use when deploys or termination may interrupt useful work.
Test representative web journeys with assistive technology and record reproducible barriers, impact, and retest evidence. Use when automated checks cannot prove real interaction, perception, or recovery.
Write a brief post-interview thank-you that reflects the actual conversation and confirms interest or a material follow-up. Use when an interview has ended and a timely, specific note is appropriate.
Examine how a feature could expose data, authority, money, or availability before it ships. Map assets and trust boundaries, rank realistic abuse cases, and verify mitigations. Use when a design handles sensitive data, privileged actions, untrusted input, or…
Reconstruct how a claim traveled through citations, summaries, and repetitions to its strongest available source. Use when a widely repeated statement may be distorted or circular.
Trace DNS resolution from client search and cache through recursive and authoritative answers, DNSSEC, delegation, and address selection. Use when names resolve differently across clients, networks, or time.
Trace one request identifier across services, queues, retries, and downstream calls to reconstruct its outcome. Use when a distributed operation failed, slowed, duplicated, or became ambiguous.
Set up a low-friction log and a weekly review that makes the data useful. Covers entry structure, capture trigger, honest gaps, summaries, and follow-up. Use when they say "I want to track X", "help me build a habit", or keep abandoning logs.
Track invoice due dates with verified terms, status, ownership, disputes, reminders, and payment or collection evidence. Use when receivables or payables need a simple reliable calendar.
Train their first machine learning model honestly. A small model on data they collected or may legitimately use, evaluated against a baseline, shipped as a demo page. Teaches what accuracy claims mean. Use when they say "I want to learn ML", "train a model",…
Transcribe an interview with accurate words, speakers, timecodes, uncertainty, and source provenance. Use when recorded speech must support editing, research, captions, quotation, or review.
Translate academic experience for industry by mapping verified research, teaching, service, and project work to target capabilities and outcomes. Use when specialist experience must be legible outside academia.
Assess a security signal, preserve evidence, and choose a proportionate next action. Use when monitoring, a vendor, a user, or an automated detector reports suspicious identity, endpoint, network, application, or data behavior.
Prioritize, classify, connect, and route a support queue using explicit customer-impact and response rules. Use when a team needs consistent ticket triage across channels, products, severities, or shifts.
Tune an operational alert using historical outcomes, service objectives, detection delay, and responder burden. Use when a notification is noisy, late, fragile, or repeatedly unactionable.
Read a first pay stub together, line by line. Explains gross pay, net pay, deductions, and the official forms used in the person's country. Use when they say "why is my paycheck smaller than I expected", ask about a deduction, or need help finding official…
Plan and verify address or contact-detail changes across important institutions and relationships. Use when moving, changing phone or email, or correcting identity records.
Upgrade a software dependency with release-note review, compatibility checks, lockfile control, targeted migration, and a tested rollback. Use when a library, framework, runtime, action, image, or package must move to a newer version without hiding behavioral…
Call a real public API from their project. Read the docs together, auth with a key kept out of git, handle errors and rate limits, and ship a feature powered by live data. Use when they want weather, scores, prices, or any outside data in an app, or ask "how…
Validate a cron expression against the exact scheduler dialect, time zone, calendar behavior, overlap, and operational intent. Use when scheduled execution timing must be predictable.
Validate a data migration through source custody, mapping rules, record and control totals, invariants, exceptions, behavior, and rollback evidence. Use when deciding whether migrated data is safe to accept.
Validate a shared component across behavior, variants, accessibility, content, themes, and consumers. Use when a new or changed design-system component needs evidence before adoption or release.