Use for writing, reviewing, debugging, or testing Python code that parses, inspects, transforms, builds, qualifies, optimizes, formats, or transpiles SQL with SQLGlot. Trigger on parse_one/parse, sqlglot expressions and scopes, dialect conversion, AST traversal, lineage, qualification, schema-aware rewrites, and unsupported translation. Do not use for executing SQL, database query tuning, regex-only text edits, ORM query construction, SQLFluff lint configuration, or generic SQL unrelated to SQLGlot.
Issue Planning and Automation prompt that generates comprehensive project plans with Epic > Feature > Story/Enabler > Test hierarchy, dependencies, priorities, and automated tracking.
Build production Apache Airflow DAGs with best practices for operators, sensors, testing, and deployment. Use when creating data pipelines, orchestrating workflows, or scheduling batch jobs.
Define or reconcile a project's stack, versioned technology guidance, NFRs, security, testing, deployment, and operations from product requirements, repository evidence, and user decisions. Use for stage 2, when a later stage invalidates a technical choice, or when the selected stack changes. Requires the q-core-contract companion. Part of the Quasar AI delivery skills.
Reasons from probabilistic penetrance, Bayesian pretest probability, and patient autonomy through three-generation pedigrees, ACMG/AMP variant criteria, ClinVar/ClinGen/gnomAD, NCCN and CPIC guidelines, and cascade-testing protocols while treating VUS over-upgraded to pathogenic, screening-versus-diagnostic confusion (NIPT vs amnio/CVS), and unaddressed psychosocial and GINA discrimination risk as first-class failure modes.
Challenge a position, decision or draft with a steerable panel of expert personas before you commit, send or ship it. Use when asked to poke holes, pressure-test, find what is missing, play devil's advocate, or prepare for a meeting. For a structured, cross-checked readiness verdict, use assess-release-readiness.
CI/CD principles — pipeline taxonomy, trigger design, supply-chain hygiene (SHA-pinned actions), OIDC cloud auth, caching (and the setup-* cache-key pitfalls that silently disable it), matrix strategy, test gates, release automation, deployment strategies, and running CI locally with act. Suggestion-mode (trade-offs over mandates). GitHub Actions recipes in RECIPES.md. Use when designing, reviewing, or scaffolding CI/CD workflows.
Verify UI behavior end-to-end by driving the running desktop app over the Chrome DevTools Protocol. Use when asked to verify, reproduce, or confirm a UI change, bug, or regression in the real app rather than in tests.
Build and verify Account registration form for Data-heavy application with Tailwind CSS v4 using a working semantic implementation, responsive RTL/LTR behavior, theme tokens, test plan and design verification.
- **Annotation:** Use `@SpringBootTest` only when loading the full application context or testing critical external integrations. Avoid integration overhead.
Diagnoses failing GitHub Actions checks on a PR or branch: lists the failures, ranks falsifiable hypotheses, confirms the root cause with log evidence, then applies and verifies the smallest authorized fix. Use when CI is red, a check or workflow run fails, or the user asks why a build, lint or test job failed.
Environment setup after project discovery. Collects API keys, guides external service configuration, writes .env, verifies connections, creates test users. Transitions project status from Pending to Ready. Use for "setup project", "configure env", "setup environment", "project setup", "get ready to build".
Assessing GraphQL API endpoints for introspection leaks, injection attacks, authorization flaws, and denial-of-service vulnerabilities during authorized security tests.
Build and verify Live example and code for Marketing landing with Bootstrap v5.3 using a working semantic implementation, responsive RTL/LTR behavior, theme tokens, test plan and design verification.
Architectural and quality code review: standards compliance, SOLID principles, testability, performance. Usar para revisar un archivo o directorio antes de mergear, o cuando se pide code review. Para bugs de correctness está el /code-review nativo de Claude Code; este revisa diseño y estándares.
Use after cpu-model-wiring to prove the CPU-enabled model is numerically correct. ORDERING (critical): on UNOPTIMIZED code do NOT gate correctness on full-model generation or a task harness (too slow — watchdog); the feasible correctness gate is the per-layer fingerprint parity vs the GPU/HF baseline on ONE short prefill. CONFIG LADDER (do not mistake the proxy for the gate): a TINY arch-faithful config is the BRING-UP proxy (op-TYPE coverage) ONLY — the AUTHORITATIVE correctness gate is FULL-MODEL DUMMY parity (ALL layers at REAL shapes, deterministic-dummy weights so NO real-weight load, MULTI-GPU reference when the full model exceeds one GPU), established BEFORE optimizing and reused to regression-guard EACH optimization; REAL weights only for the dtype-bridge + task finale. Optimize AFTER correctness; regression-guard with the SAME fast parity; full generation + task accuracy come LAST once fast. LAYERED checks so a failure localizes: (0) a cheap real-prompt COHERENCE smoke test first ('it ran' on random/dummy inputs is NOT 'it's correct'; catches gross zeroed/no-op/routing/layout bugs and silently-stubbed forward paths), (0.5) a per-kernel LOW-BIT PARITY gate for every dtype bridge (fp4/mxfp4/fp8/int4 weights on a bf16/fp8 kernel) — kernel output vs an independent torch dequant oracle from the same packed bytes, because 'a bf16/fp8 kernel exists' is coverage, not correctness, (1) per-layer / logits parity vs a trusted reference (HF or the GPU SGLang path) under tolerance, and (2) end-to-end task accuracy via the in-repo harnesses (gsm8k, mmlu, hellaswag) within N points. Includes a correct-by-construction ISOLATION technique to bisect MoE vs shared-path bugs. Blocks enablement on any regression; a per-layer diff pinpoints the offending op for the wiring step to fix.