▸case-01 I'm organizing the developer documentation for our backend services. We have permanent project-wide architectural rules as well as a detailed step-by-step checklist for database migrations. How should I split these between our main project instruction files and standalone skill files so the main configs don't get bloated? | fail→pass | 14,521 | 9,837 | -32% | 1 | 1 | 0% | 2,310 | 1,878 | -19% | 0 | 0 | — |
▸case-02 Our engineering team is writing identical logic for event-streaming and third-party API integration across several backend microservices. Should I create a standalone backend skill for this pattern, and if so, what standard sections need to be included to make the guide actionable? | fail→pass | 15,301 | 10,824 | -29% | 1 | 1 | 0% | 2,270 | 1,966 | -13% | 0 | 0 | — |
▸case-03 I'm writing an AGENTS.md file inside a specific sub-module of our backend services. How should I link it to existing shared skills like user authentication flows or authoring flows, and where should focused skills be located? | fail→pass | 13,053 | 6,088 | -53% | 1 | 1 | 0% | 2,124 | 1,304 | -39% | 0 | 0 | — |
▸case-04 We are configuring repository guidelines for our backend project. The team is debating whether to put compiler settings, directory structure invariants, and reusable step-by-step API integration playbooks all together in root CLAUDE.md or split them. Provide a clear file placement guide. | fail→pass | 14,868 | 12,036 | -19% | 1 | 1 | 0% | 2,358 | 2,157 | -9% | 0 | 0 | — |
▸case-05 I am drafting repository guidelines for a backend team. The team wants to store a complex 12-step emergency database recovery procedure directly inside CLAUDE.md alongside project-wide naming conventions. What is the proper file split to keep root configuration files concise? | fail→pass | 11,419 | 6,605 | -42% | 1 | 1 | 0% | 1,853 | 1,205 | -35% | 0 | 0 | — |
▸case-06 Our backend team handles OAuth authentication across four separate service modules, using identical refresh token logic in each. A teammate suggests leaving this code undocumented or putting a copy-pasted guide in each folder. What is the recommended strategy for managing this repeated authentication pattern in the repository? | fail→pass | 12,058 | 11,145 | -8% | 1 | 1 | 0% | 1,823 | 1,477 | -19% | 0 | 0 | — |
▸case-07 We have multiple backend services using the same MongoDB persistence and retry logic across multiple endpoints. A developer proposed creating a generic 12-page Wiki article outside the codebase. How should this repeated persistence workflow be documented and structured within the project repo? | fail→pass | 15,507 | 10,876 | -30% | 1 | 1 | 0% | 2,502 | 1,989 | -21% | 0 | 0 | — |
▸case-08 I am refactoring a single backend utility function that formats timestamps for a legacy reporting endpoint. This task is a one-off edit and will not be repeated. Should I create a new standalone backend skill file in skills/ for this task, or handle it directly in the codebase? | pass→pass | 7,383 | 3,090 | -58% | 1 | 1 | 0% | 1,059 | 734 | -31% | 0 | 0 | — |
▸case-09 Our backend service has a module-specific payment handling workflow that only applies to the payment sub-service. Where in the repository directory hierarchy should focused backend module skills be placed so sub-module instruction files can point to them? | fail→pass | 13,182 | 4,624 | -65% | 1 | 1 | 0% | 2,171 | 937 | -57% | 0 | 0 | — |
▸case-19 Our team repeats the exact same Kafka event streaming publication pattern across 5 microservices. Developers are currently re-implementing setup and validation logic from scratch every time. What strategy should we use to make this backend workflow reusable and actionable? | fail→pass | 15,382 | 10,200 | -34% | 1 | 1 | 0% | 2,497 | 1,806 | -28% | 0 | 0 | — |
▸case-10 I need to author a backend skill document for a complex debugging playbook. I am tempted to include 8 sections: overview, architecture, prerequisite tools, environment setup, execution steps, troubleshooting, performance tips, and appendix. What minimal set of 4 core actionable sections must be present in every skill document? | fail→pass | 14,093 | 3,325 | -76% | 1 | 1 | 0% | 2,202 | 757 | -66% | 0 | 0 | — |
▸case-11 When writing a standalone SKILL.md file for a zero-downtime database migration checklist, what specific section names must be included to ensure the guide is structured and actionable? | fail→pass | 27,445 | 3,274 | -88% | 1 | 1 | 0% | 1,026 | 710 | -31% | 0 | 0 | — |
▸case-12 In our backend repository index, what exact path relative to the repository root is assigned to the shared authoring flow skill file? | fail→pass | 5,476 | 4,484 | -18% | 1 | 1 | 0% | 757 | 963 | +27% | 0 | 0 | — |
▸case-13 What is the exact repository relative path for the shared backend user authentication skill file listed in the backend index? | fail→fail | 9,168 | 2,865 | -69% | 1 | 1 | 0% | 1,409 | 619 | -56% | 0 | 0 | — |
▸case-14 I am writing a sub-module instruction file `src/api/orders/AGENTS.md`. I want to reference the shared user auth workflow skill and also create a local skill for order-specific persistence workflows. What path structure should I use to reference these? | fail→fail | 9,731 | 6,404 | -34% | 1 | 1 | 0% | 1,719 | 1,437 | -16% | 0 | 0 | — |
▸case-15 I need to implement a REST endpoint in Node.js/Express that takes a user payload, validates schema using Zod, and returns HTTP 201. Should I write a custom AGENTS.md or SKILL.md file for this single REST endpoint handler, or write standard Express route handler code? | pass→fail | 9,990 | 7,176 | -28% | 1 | 1 | 0% | 1,888 | 1,300 | -31% | 0 | 0 | — |
▸case-16 We are designing a PostgreSQL schema for indexing user session tokens to speed up lookups. What database indexing strategy (e.g. B-Tree on user_id and session_token) should we apply for fast lookup performance? | pass→pass | 18,101 | 15,708 | -13% | 1 | 1 | 0% | 2,934 | 2,817 | -4% | 0 | 0 | — |
▸case-17 Our team has strict architectural rules regarding layer dependencies (e.g., controllers cannot import ORM models directly). Should these durable, invariant rules be placed in individual SKILL.md files or in directory-level instruction files like AGENTS.md and CLAUDE.md? | pass→pass | 12,885 | 3,969 | -69% | 1 | 1 | 0% | 1,962 | 862 | -56% | 0 | 0 | — |
▸case-18 We are writing a deployment rollback playbook that steps developers through verifying cache invalidation and database rollback steps. To prevent project instruction files like CLAUDE.md from becoming overcrowded, where should this operational checklist be placed? | fail→pass | 8,985 | 4,741 | -47% | 1 | 1 | 0% | 1,526 | 1,048 | -31% | 0 | 0 | — |
▸case-20 When documenting a repeated database persistence workflow in a backend repository, what standard actionable sections must be present in the skill file? | fail→pass | 15,205 | 3,358 | -78% | 1 | 1 | 0% | 2,369 | 657 | -72% | 0 | 0 | — |
▸case-21 We are building an inventory of shared skills for our backend API repo. Which two specific backend skill paths are already established under the `skills/` directory for authoring and authentication? | fail→pass | 6,924 | 1,803 | -74% | 1 | 1 | 0% | 1,119 | 461 | -59% | 0 | 0 | — |
▸case-22 In a backend service, we want to know when it is appropriate to create a new backend skill document. What specific conditions justify adding a focused skill file instead of just putting code in the sub-module? | fail→pass | 13,619 | 11,021 | -19% | 1 | 1 | 0% | 2,050 | 1,952 | -5% | 0 | 0 | — |