LumabyteCo/clarifyprompt-mcp

What's new in 1.8.0

Step #3 of the MCP modernization roadmap: the engine's read surfaces become browseable resource templates with argument autocompletion. No tool or engine behavior changes — purely additive, read-only.

Resource templates

Four templates join the static clarifyprompt://categories, each with list + complete + read, backed by an existing engine getter:

| URI template | What it reads |
|---|---|
| clarifyprompt://platforms/{category}/{id} | One platform's full config — resources/list enumerates all 60+ as individual URIs |
| clarifyprompt://traces/{date} | Optimization-trace summary index for a UTC day |
| clarifyprompt://packs/{id} | One loaded knowledge pack's metadata |
| clarifyprompt://memory/facts/{scope} | Live remembered facts under a scope |

MCP hosts with a resource browser (Claude Desktop, Cursor) now get a navigable tree instead of a single static blob.

Autocomplete

completion/complete resolves the template variables — {category} → the 7 category ids, {id} → platform ids scoped by the chosen {category} (e.g. "mid" + category=image → midjourney), {date} → days with traces, pack ids, memory scopes.

MCP's completion applies to prompt arguments and resource-template variables only — not tool inputs (the spec has no tool-argument completion). ClarifyPrompt registers no prompts, so autocomplete lives entirely on the resource templates.

Capabilities

The server now advertises resources (with templates) and completions at initialize.

Tests

New deterministic npm run test:resources battery (no LLM/embeddings) locks the surface: capabilities advertised, all four templates list, pure-registry reads, completion resolution, graceful empty-day trace reads. Added to test:all.

Verified

test:resources · wire · integration 9/9 · evals 26/26 on gemma4:31b-cloud (0 output-validation errors) · all CI build/docker/secrets jobs · evals on gpt-4o-mini.

Install

```bash
npm install -g [email protected]
```