▸case-05 We found an existing OpenAPI generator skill in our repository that covers 80% of our requirements for REST API documentation. I am planning to create a brand new duplicate skill from scratch named rest-doc-generator-v2 to avoid touching the existing skill. Research this scenario and recommend how we should proceed. | pass→pass | 24,610 | 12,960 | -47% | 1 | 1 | 0% | 2,234 | 1,879 | -16% | 0 | 0 | — |
▸case-01 I need to add a PDF parsing capability to my AI setup. Can you look into existing skills or community examples to see if there is already a good solution, analyze what patterns work best, and advise me on whether I should install something pre-made or build a custom one? | fail→fail | 17,430 | 17,861 | +2% | 1 | 1 | 0% | 2,630 | 1,928 | -27% | 0 | 0 | — |
▸case-02 We are looking for agent skills that handle database migrations. Please check available repositories and local skill configs for existing tools, point out any feature gaps between them, and summarize the best design practices I should follow before writing our own. | fail→pass | 23,101 | 11,835 | -49% | 1 | 1 | 0% | 3,123 | 2,080 | -33% | 0 | 0 | — |
▸case-03 Can you research available skills for automated code reviewing? I want a breakdown of existing implementations, an assessment of common design patterns, and a final recommendation on whether to adopt an existing tool or create a brand new specification. | fail→fail | 47,614 | 19,035 | -60% | 1 | 1 | 0% | 6,567 | 2,908 | -56% | 0 | 0 | — |
▸case-04 I am searching for existing agent skills that handle legacy COBOL mainframe JCL file parsing. Check the repositories for matching skills. If none are found, I'm tempted to just give up and tell the team to write raw Python scripts outside the skill architecture. Give me an assessment and next steps. | fail→pass | 17,348 | 15,948 | -8% | 1 | 1 | 0% | 2,845 | 2,412 | -15% | 0 | 0 | — |
▸case-06 Our team wants to research agent capabilities for cloud infrastructure management on AWS. The developer wants to manually browse web search results and copy-paste random GitHub links. I want a structured skill research workflow that leverages automated discovery tools for finding skills. Outline the search step using relevant integration engines. | fail→pass | 23,320 | 9,663 | -59% | 1 | 1 | 0% | 2,849 | 1,811 | -36% | 0 | 0 | — |
▸case-07 I have a rough plain-English description of a skill for automated Slack notifications on deployment failures. I'm thinking of manually mapping out all parameters in a spreadsheet first before looking at skill ecosystems. Can you analyze existing skill examples and show how natural language descriptions can be transformed into structured skill inputs? | fail→fail | 31,358 | 11,977 | -62% | 1 | 1 | 0% | 3,320 | 2,291 | -31% | 0 | 0 | — |
▸case-08 We researched existing Kubernetes troubleshooting skills and found several gaps, so we need to author a custom path. My colleague suggests writing custom bash scripts directly in CI/CD pipelines instead of leveraging skill creation tools. Provide the workflow step for setting up the custom authoring path using appropriate skill creation integrations. | pass→fail | 26,158 | 9,023 | -66% | 1 | 1 | 0% | 2,728 | 1,792 | -34% | 0 | 0 | — |
▸case-09 I need to find skill examples for CSV data validation. The team believes we should only query online web APIs and completely ignore local environment configurations like local dotfiles or local skill hidden folders. Evaluate existing skills and advise on where to look locally. | fail→pass | 25,689 | 10,450 | -59% | 1 | 1 | 0% | 3,019 | 2,072 | -31% | 0 | 0 | — |
▸case-10 We need to research skills for Docker container log analysis. I'm tempted to immediately write a custom Python script from scratch without checking repositories, comparing gaps, or summarizing patterns. Walk me through the required research process and make a final recommendation. | fail→pass | 15,545 | 11,661 | -25% | 1 | 1 | 0% | 2,524 | 2,252 | -11% | 0 | 0 | — |
▸case-11 I am evaluating skills for GraphQL schema generation. Someone suggested we just pick the first skill found in agentskill.sh without analyzing missing capabilities or limitations compared to our needs. Perform a skill research review focusing on gap identification before recommending action. | pass→pass | 20,236 | 13,210 | -35% | 1 | 1 | 0% | 2,736 | 2,298 | -16% | 0 | 0 | — |
▸case-12 We are looking at multiple existing skills for Terraform state auditing in .grok/skills/. A team member suggests ignoring the common architectural patterns in these skills and inventing a completely custom framework paradigm. Conduct a skill evaluation that extracts the best practices to adopt. | fail→pass | 20,563 | 13,191 | -36% | 1 | 1 | 0% | 2,978 | 2,267 | -24% | 0 | 0 | — |
▸case-13 Search for existing agent skills handling Qiskit quantum circuit optimization. If no skills exist in local or remote registries, my manager wants to abandon the agent skill model and use legacy C++ binaries. Provide the proper error handling response when no matches are found. | fail→pass | 11,868 | 9,214 | -22% | 1 | 1 | 0% | 1,221 | 1,453 | +19% | 0 | 0 | — |
▸case-14 We found a git-pr-labeler skill on agentskill.sh that performs basic automated labeling on PR titles, but lacks regex body parsing. I want to build an entirely separate PR labeling skill named pr-body-labeler-v1 from scratch. Research this setup and tell me the recommended path. | fail→fail | 22,195 | 11,654 | -47% | 1 | 1 | 0% | 3,345 | 2,385 | -29% | 0 | 0 | — |
▸case-15 I am looking for skills that handle Redis cache purging. A developer claims we only need to check remote URLs and never look at .grok/skills/ on the local system. Provide a complete search scope for evaluating existing cache skills. | fail→pass | 15,922 | 10,561 | -34% | 1 | 1 | 0% | 2,427 | 2,014 | -17% | 0 | 0 | — |
▸case-16 We are researching skills for Markdown table formatting. Should I recommend whether to install or custom author a skill before searching agentskill.sh and local .grok/skills/? Outline the correct sequence of steps for researching skills. | fail→pass | 9,463 | 5,380 | -43% | 1 | 1 | 0% | 1,600 | 1,128 | -30% | 0 | 0 | — |
▸case-17 Our team leader said 'find skill examples' for SQL query optimization. I am tempted to search StackOverflow for generic Python functions. Execute the skill research workflow to find relevant skill examples and establish best practices. | fail→pass | 20,961 | 11,339 | -46% | 1 | 1 | 0% | 2,902 | 2,104 | -27% | 0 | 0 | — |
▸case-18 A junior dev asked for 'skill best practices' when designing CLI interaction skills. I'm tempted to hand them a general Python PEP-8 style guide. Conduct a research review of existing CLI skills to identify true skill design patterns. | fail→fail | 24,710 | 13,079 | -47% | 1 | 1 | 0% | 3,712 | 2,109 | -43% | 0 | 0 | — |
▸case-19 My manager instructed me to 'research skills' for Kafka event streaming integrations. I'm considering writing a lengthy research paper on Kafka architecture instead of looking for agent skills. Provide the structured skill research analysis. | fail→pass | 17,401 | 11,746 | -32% | 1 | 1 | 0% | 2,735 | 2,003 | -27% | 0 | 0 | — |
▸case-20 Write a complete SKILL.md file for a Kubernetes pod log analyzer skill. Include triggers, workflow steps, and error handling tables directly in your response. | pass→pass | 18,311 | 14,068 | -23% | 1 | 1 | 0% | 3,168 | 2,481 | -22% | 0 | 0 | — |
▸case-21 Run the local docker-container-cleaner skill on my system right now to remove exited containers and purge dangling images. | pass→pass | 3,651 | 8,702 | +138% | 1 | 1 | 0% | 583 | 1,543 | +165% | 0 | 0 | — |
▸case-22 I am getting a Python KeyError: 'headers' in line 45 of my local http-request-skill/main.py file. Here is the stack trace. Please debug this code error and provide the fixed Python snippet. | pass→pass | 8,362 | 8,193 | -2% | 1 | 1 | 0% | 1,392 | 1,674 | +20% | 0 | 0 | — |
▸case-23 We need to set up a pyproject.toml file to manage poetry dependencies for our skill development repository. List the required dependencies and lockfile configurations. | pass→pass | 16,874 | 12,056 | -29% | 1 | 1 | 0% | 2,570 | 2,325 | -10% | 0 | 0 | — |