# Observability And Research

> Combined observability check with research discovery. Validates Langfuse is healthy while concurrently searching for papers on agent observability.

- Skill: `knuckles-team/observability-and-research` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add knuckles-team/observability-and-research`
- Raw SKILL.md: https://api.skillmd.com/api/skills/knuckles-team/observability-and-research/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: Knuckles-Team (https://skillmd.com/u/knuckles-team)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/knuckles-team/observability-and-research

---


# Observability And Research Workflow

**CONCEPT:RESEARCH-001**

Combined observability check with research discovery. Validates Langfuse is healthy while concurrently searching for papers on agent observability.

## Steps

### Step 0: Langfuse Configuration Inventory [skill: langfuse-mcp]
**Agent**: `search-agent`
**Tools**: `sx_search, graph_query`

Check Langfuse health, list all score configs, and list all datasets
Expected: `health, score, dataset`

### Step 1: Scholarx Mcp
**Agent**: `analyzer-agent`
**Tools**: `graph_analyze, sx_storage`

Search for recent papers on LLM observability and agent tracing
Expected: `paper, observability`

### Step 2: Langfuse Project Inventory [skill: langfuse-mcp]
**Agent**: `synthesizer-agent`
**Tools**: `graph_analyze, document_tools`

List all current projects in the Langfuse instance
Expected: `project`

### Step 3: KG Persistence [depends_on: Step 2]
**Agent**: `synthesizer-agent`
**Tools**: `graph_write`

Persist workflow results as nodes and edges in the Knowledge Graph.
Create appropriate typed nodes with metadata and link to existing domain entities.

## Output
- Observability And Research results persisted in KG
- Structured report (MD/PDF)
- Audit trail with timestamps and agent attributions

## Execution

Run this workflow as a dependency-ordered DAG. Steps with no unmet `depends_on` run in parallel; dependents run after their prerequisites complete.

- **Run first (in parallel):** Step 0 — Langfuse Configuration Inventory; Step 1 — Scholarx Mcp; Step 2 — Langfuse Project Inventory
- **After level 0:** Step 3 — KG Persistence

**Execution:** If graph-os is reachable, offload the whole DAG via `graph_orchestrate action=execute_workflow` (or the `kg-delegate` skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet `depends_on` in parallel, then their dependents.