From grepai
Use when GrepAI seems broken or you need to verify all components are running correctly
npx claudepluginhub jugrajsingh/skillgarden --plugin grepaiThis skill is limited to using the following tools:
Diagnose all GrepAI components and report status with indicators.
Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.
Guides building MCP servers enabling LLMs to interact with external services via tools. Covers best practices, TypeScript/Node (MCP SDK), Python (FastMCP).
Generates original PNG/PDF visual art via design philosophy manifestos for posters, graphics, and static designs on user request.
Diagnose all GrepAI components and report status with indicators.
Run each check sequentially. Collect results, then print a unified report.
Check by image ancestry rather than hardcoded container names:
docker ps --filter ancestor=ollama/ollama --format "{{.Names}}\t{{.Status}}\t{{.Ports}}"
docker ps --filter ancestor=qdrant/qdrant --format "{{.Names}}\t{{.Status}}\t{{.Ports}}"
docker ps --filter ancestor=postgres --format "{{.Names}}\t{{.Status}}\t{{.Ports}}"
Record: which services are running.
curl -s --max-time 5 http://localhost:11434/api/tags
Record: reachable? Parse response for available models.
From the /api/tags response, list models that are embedding models:
Record: which embedding models are pulled.
Check if a postgres container is running (from step 1). If so, get the container name and check:
docker exec {POSTGRES_CONTAINER} pg_isready -U grepai
Skip if config shows GOB or qdrant backend and no postgres container is running. Record: accepting connections?
docker exec {POSTGRES_CONTAINER} psql -U grepai -d grepai -tAc "SELECT extname FROM pg_extension WHERE extname='vector'"
Skip if config shows GOB or qdrant backend. Record: vector extension installed?
Check Qdrant REST API on port 6333 (not 6334, which is gRPC):
curl -s --max-time 5 http://localhost:6333/collections
Skip if config does not use qdrant backend and no qdrant container is running. Record: reachable? Collection count.
Read: .grepai/config.yaml
If missing, report not initialized. Otherwise extract and report:
grepai status
Record: indexed files, chunks, last updated.
claude mcp list
Or read ~/.claude/mcp.json and ./.claude/mcp.json. Record: grepai MCP server registered?
grepai watch --status
Record: running or not.
grepai workspace list
If workspaces exist, get details for each:
grepai workspace show {NAME}
Optionally also read ~/.grepai/workspace.yaml for additional detail.
Record: configured workspaces, backend type, project count per workspace. Skip if no workspaces configured.
grepai workspace status {NAME}
And check workspace watcher:
grepai watch --workspace {NAME} --status
Record: per-workspace indexing health and watcher state. Skip if no workspaces configured.
Print unified status report with OK/FAIL/WARN indicators for each component. If any show FAIL, print targeted fix suggestions.
See references/report-format.md for the report template and troubleshooting hints table.