cursor-develop-autorules

Test API endpoints with proper authorization including Basic Auth and session cookies. Use when testing curl requests, checking API responses, getting 401 Unauthorized, session expired errors, or verifying endpoint behavior after changes.

Implement accumulated improvements from improvements-backlog into AI rules and skills. Use when backlog has 5+ items, weekly rule update cycle, 2+ high priority improvements exist, before important releases, when same error repeated 3+ times, when user says implement improvements/process backlog/update rules from errors.

Fix bugs, errors, exceptions, and unexpected behavior. Use when debugging, fixing errors, handling crashes, or when user mentions bug, error, broken, not working, crash, exception, 500 error, Application error, white screen, TypeError, undefined is not a function, or unexpected behavior.

Analyze Claude Code prompt cache efficiency and costs. Use when user asks about cache performance, token usage, session costs, how much a session cost, wants to optimize prompt caching, when costs seem high, when context window is filling up, or mentions cache-analyzer/cache hits/token costs/session metrics/cost analysis.

Review code for quality, security, and maintainability. Use when reviewing pull requests, examining code changes, doing QA checks, architecture review, or CTO review. Also trigger for COMPLEX tasks affecting more than 5 files, API contract changes, or data migrations.

Create new skills, modify and improve existing skills, create or update AI rules. Use when creating rules, building skills, adding coding standards, updating .claude/rules/ or .claude/skills/, converting Cursor rules to Claude Code, measuring skill performance, running evals, or optimizing skill descriptions.

Deploy backend and/or frontend services. Use after npm run build completes, when restarting services, when user mentions deploy/build/restart, or when page shows Application error. Also trigger on 502/503 errors, service health check failures, or after code changes that need deployment.

Develop new features, functionality, and enhancements. Use when creating, adding, building, implementing new features, components, endpoints, pages, or modules. Also trigger when user asks to add functionality, create a form, implement a filter, build a new UI section, or add an API endpoint.

Analyze and fix issues in the last completed task. Use when user reports errors after task was marked as done, when completion report had high confidence but result has issues, when AI missed obvious problems, or when post-task issues appear.

Find gaps, stubs, broken workflows, and incomplete features in code. Use before implementing features, for code audits, when checking TODO coverage, or when searching for buttons without handlers, forms without submit, links to nowhere, or partial CRUD implementations.

Improve code structure without changing behavior. Use when refactoring, cleaning up, simplifying, extracting methods, reducing complexity, removing duplication, or applying DRY principle. Also trigger when files exceed size limits or code smell is detected.

Analyze data, investigate datasets, debug with data, and explore system behavior. Use when analyzing, running SQL queries, data profiling, investigating patterns, building statistics, exploring CSV/JSON data, or debugging with data evidence.

Review work session quality and capture improvement opportunities. Use at end of session, after completing complex features, after series of errors or failed attempts, after 2+ hours of focused work, when user says done/finished/wrap up, for retrospective analysis, or when same mistake happened twice in a session.

Test-Driven Development with strict Red-Green-Refactor cycle. Use when developing with TDD, writing tests before code, adding test coverage, or when user mentions test-first, red-green-refactor, or wants reliable test suites.

Scan codebase for technical debt including oversized files, duplicated code, code smells, and TODO items. Use when files exceed size limits, when TODO/FIXME/HACK count is growing, when build is slow, when coverage is low, when user says cleanup/tech debt/code quality, or when empty catch blocks or dead code are suspected.

Внедрение накопленных улучшений из improvements-backlog в инструкции AI агента

Test API endpoints with proper authorization. Use when testing curl requests, checking API responses, or getting 401 Unauthorized. API requires two auth levels - Basic Auth (nginx) + Session Auth (login). Credentials in .cursor/.secrets/.

Deploy backend and/or frontend after code changes. Use after npm run build, when deploying, restarting services, or when user mentions deploy, build, restart, or sees "Application error". BUILD WITHOUT RESTART = BROKEN SITE.

Develop new features, functionality, and enhancements. Use when creating new features, adding functionality, building components, or when user says add, create, build, develop, implement. Includes Duplicate Check, JTBD Analysis, TDD approach, UI pattern reuse, and service restart verification.

Analyze and fix issues in the last completed task — find missed problems, fix errors, perform session review. Use when user reports errors after task was marked as done, when completion report had high confidence but result has issues, or when AI missed obvious problems.

Find gaps, stubs, broken workflows, and incomplete features in code. Use before implementing features (to find implicit requirements) or for code audits (to find TODO, FIXME, empty handlers, broken workflows). Keywords - gap analysis, audit, find stubs, incomplete, broken workflow.

Analyze data, investigate datasets, work with CSV/parquet/pandas/dataframes. Use when analyzing data, exploring datasets, running experiments, or when user mentions data, analysis, parquet, csv, pandas, dataframe, statistics, investigation.

Review work session quality and capture improvements. Use at end of session, after large tasks, after series of errors, or when user asks for session review, retrospective, lessons learned. Records improvements to backlog.

Scan codebase for technical debt and fix safely with TDD. Use to find oversized files, duplicated code, code smells, and refactor safely. Workflow - SCAN, TEST CASES, REFACTOR, VERIFY. Keywords - techdebt, tech debt, duplicates, code quality audit.

Fix bugs, errors, exceptions, and unexpected behavior. Use when debugging, fixing errors, handling crashes, or when user mentions bug, error, broken, not working, crash, exception, 500 error, Application error. Uses 5 Whys Root Cause Analysis.

Review code for quality, security, and maintainability. Use when reviewing pull requests, examining code changes, doing code review, QA checks, or architecture review for COMPLEX tasks. Includes QA checklist and CTO review template.

Create or update Cursor Rules (.mdc files) and Skills (SKILL.md). Use when creating rules, adding coding standards, setting up conventions, updating .cursor/rules/, or converting rules to skills. Defines standard format, naming, frontmatter, token budget.

Improve code structure without changing behavior. Use when refactoring, cleaning up, optimizing, simplifying, extracting methods, or reducing complexity. Requires tests BEFORE refactoring, small steps, commit after each green test.

Test-Driven Development workflow with strict Red-Green-Refactor cycle. Use when developing features with TDD, writing tests before code, or when test-driven approach is needed. MANDATORY order - test cases table BEFORE code, failing tests BEFORE implementation.