AFTER

Use for tasks that define, repair, validate, or preserve external interface contracts: REST or FastAPI endpoints, Pydantic models, JSON Schema, OpenAI tool/function schemas, ChatML or SFT JSONL formats, request batching, streaming callbacks, and client-visible compatibility behavior.

Use for configuration repair and validation tasks involving YAML, JSON, TOML, environment variables, Nginx, Docker Compose, Kubernetes manifests, feature flags, service settings, logging, rate limits, and runtime behavior controlled by config files.

Author Dockerfiles, docker-compose stacks, and Kubernetes manifests as text artifacts: multi-stage builds, layer caching, image-size discipline, service composition with health checks, Deployment/Service/ConfigMap manifests, and pure-Python verification of the resulting YAML/Dockerfile.

Use for bug-fixing and failure-investigation tasks: reproduce failing tests or builds, trace dependency or configuration errors, patch minimal code, audit vulnerabilities, fix package/build breakage, and verify the repaired behavior with regression tests.

Create, inspect, and fill Word `.docx` files with python-docx, including robust template replacement across paragraphs, tables, nested tables, headers, and footers.

Use for tasks that compute, compare, optimize, or report metrics: classification and retrieval metrics, campaign uplift, model or prompt comparisons, LLM-as-judge reports, statistical significance, budgeted optimization, and behavior-preserving performance improvements.

Use for evidence and safety verification tasks: detect prompt injection, mask PII, classify factual or policy-sensitive claims, redact sensitive text, compare claims to trusted evidence, and produce auditable reports with precision, recall, F1, false-positive, or leakage checks.

Use for migration tasks that convert systems or artifacts while preserving behavior: SQL schema/data migrations, Terraform or IaC migrations, serialization format migrations, rollback planning, state movement, compatibility checks, and before/after validation.

Use for tasks that train, fine-tune, adapt, or evaluate predictive models: AutoML tabular models, time-series forecasting, custom PyTorch modules, object detection, QA fine-tuning, LoRA/support models, checkpoints, predictions, and metrics files.

Use for tasks where a PDF file is the primary input or output: extract text or tables, fill AcroForm fields, redact sensitive or authorship-leaking content, generate PDFs from structured records, inspect page-level statistics, or prove that a produced PDF is readable and non-recoverable where redaction is required.

Use for data, ML, CI, and monitoring pipelines where the task is to assemble inputs, transform records, preserve dependencies, process streams, optimize workflow execution, or produce durable outputs and run statistics.

Use for PowerPoint tasks where a .pptx deck is read, created, edited, reformatted, compared, or validated, including slide layouts, placeholders, theme styles, speaker notes, images, charts, and reference-format matching.

Cheatsheet of prompt-engineering techniques: zero/few-shot, chain-of-thought, self-consistency, ReAct, tree-of-thought, role and system prompts, JSON-schema outputs, tool/function definitions, decomposition, critique-and-revise, structured extraction, injection defenses, and rubric evaluation.

Use for retrieval and grounding tasks: build or debug BM25, dense, hybrid, FAISS, metadata-filtered, or RRF retrieval pipelines; rank documents for queries; evaluate NDCG/MAP/Recall; and produce grounded outputs with document IDs, scores, ranks, and source traces.

Use for code-quality refactoring tasks: extract function from a long method, rename a confusingly-named symbol, restructure a tangled module, replace a conditional dispatch with polymorphism, split a god-class along its responsibilities, and introduce dependency injection so I/O can be faked in tests. Behavior must be preserved; tests stay green; public API stays intact unless the task says otherwise.

Reference for writing and tuning SQL on tabular sources: SELECT/INSERT/UPDATE/DELETE, INNER/LEFT/RIGHT/FULL/SEMI/ANTI joins, window functions (ROW_NUMBER, RANK, LAG/LEAD, running aggregates), recursive and non-recursive CTEs, EXPLAIN/EXPLAIN ANALYZE, index design, and the three Python access layers (DuckDB embedded analytics over Parquet/CSV, SQLAlchemy core/ORM for multi-dialect modeling, psycopg2 for direct PostgreSQL with cursor control). Includes worked feature-extraction and slow-query-diagnosis examples.

Reference for applied statistics on tabular data: descriptives, confidence intervals, one/two-sample and paired tests (t-test, Mann-Whitney, Wilcoxon, chi-square, Fisher exact, ANOVA, Kruskal-Wallis), assumption checks (Shapiro, Levene, Bartlett), correlation (Pearson, Spearman, Kendall), effect sizes (Cohen's d, Cliff's delta, eta-squared, phi), OLS with residual diagnostics, multiple-comparison correction (Bonferroni, BH), and a single power-analysis example via statsmodels.

Use for authoring Terraform / HCL infrastructure-as-code: provider declarations (AWS, GCP, Azure), resources (VPC, subnet, security group, EC2, S3 / GCS, IAM, RDS), variables, locals, outputs, modules (calling and authoring), data sources, count vs for_each, lifecycle meta-arguments (prevent_destroy, create_before_destroy, ignore_changes), backends and remote state, and workspaces.

Author Python test suites graders execute: pytest unit tests with fixtures and tmp_path, parametrized edge cases, mocking external boundaries with unittest.mock, property-based invariants via hypothesis, integration markers with conditional skips, coverage measurement with thresholds, and unittest.TestCase for legacy suites.

Use for tasks over transactional / event data: aggregating raw event logs into entity-level features (RFM, time-window stats), per-entity sequence representations (CoLES embeddings or hand-crafted lag features), double-entry ledger reconciliation and balance audits, point-of-sale category-share / basket analysis, and per-transaction anomaly / fraud scoring.

Use for tasks that check data, artifacts, citations, schemas, or cross-format records against explicit rules and produce clean outputs, violation lists, anomaly reports, reconciliation diffs, or validation summaries.

Use for spreadsheet tasks where the workbook itself matters: read, repair, populate, reconcile, calculate, format, or export XLSX/XLSM/CSV/TSV artifacts while preserving sheets, formulas, number formats, styles, tables, named ranges, and deterministic cell-level validation.