Routes a described situation to the right skills in the right order.
npx skills add https://github.com/dineshrevunuru/SuperSkills --skill ask-dinesh-dineshrevunuru
Please help me install this Agent Skill. Skill: ask-dinesh Source: https://github.com/dineshrevunuru/SuperSkills/tree/main/ask-dinesh Command: npx skills add https://github.com/dineshrevunuru/SuperSkills --skill ask-dinesh-dineshrevunuru
High-intent search queries and answers about installing and using this skill.
Describe your situation — the object, the stage of work, and the stakes — and the router returns the matching skill or an ordered chain. It reads the live skill map and confirms each skill exists on disk before naming it.
It runs a routing gate that infers three facts: the object (screen, study, code, written piece), the stage (framing, researching, designing, building, reviewing), and the stakes (scrappy versus external-facing). It asks only the facts it cannot infer.
Skip the router when the fit is obvious — a one-skill ask like writing a survey should load that skill directly. Routing a single-skill request through the full decision tree adds overhead with no benefit.
It does not. The router verifies every candidate skill against the directory listing before naming it, and if a needed skill is genuinely unbuilt, it routes to the nearest real skill and points at the roadmap instead of inventing one.
No. It outputs the ordered skill chain, names the underlays that load beneath it, marks which steps are cuttable, and stops. Each routed skill owns its own method once named.
aydadevelopSummarize project progress and route to the next action.
ImaginerLabsRefine LLM-generated content using structured elicitation methods from CSV.
AarononthewebGenerate routing indexes mapping tasks to dotnet-skills in AGENTS.md and CLAUDE.md.
tabgabRoute opp_repl user tasks to the exact sibling skills via a decision tree.
