Coding Agent support tools, knowledge management, guardrails, constraints
Command Claude Code
Check code or an action against this project's constraint rules, read compliance status and violation history, and add or amend rules. Use before writing code that might trip a guardrail, when a PreToolUse hook has blocked something and you need to see why, or when the user asks about constraints, guardrails or…
Command Claude Code
Enforce consistent styling for documentation artifacts (PlantUML, Mermaid, markdown, PNG diagrams).
Command Claude Code
Describe a cross-agent experiment in plain English (or with flags), then auto-run the matrix, compare, and render the ranked variant table.
Command Claude Code
Query and rebuild the project's code knowledge graph (graphify). Use for any question about the codebase — architecture, "what calls X", "where is Y", file/function relationships, data-flow tracing — instead of blind greps. The graph is a static graph.json served over MCP by the coding-graphify container; rebuild it…
Command Claude Code
Run, resume, regrade and report the kgbench code-retrieval benchmark (coding-v1) across retrieval arms, agents and models — and diagnose the routing that decides which model actually answers. Also covers the Performance → Benchmarks dashboard sub-tab, the second front-end over the same scripts.
Command Claude Code
Use this skill whenever the user wants to automate a browser, scrape web content, take screenshots or PDFs of pages, fill out forms, click through UI flows, or run end-to-end tests — without using an MCP server. This skill drives Playwright directly from the bashtool via Node.js scripts. Trigger whenever the user says…
Command Claude Code
Run UKB knowledge-base workflows (wave-analysis and friends), manage the ontology, refresh stale knowledge entities, and reach the semantic-analysis tools. Use whenever the user says "ukb", "ukb full", "ukb debug", asks to run or check a knowledge workflow, or asks about ontology classes and entity refresh. Replaces…
Command Claude Code
Load session logs (LSL) from current and coding projects for continuity.