From Skills Back to Tools: Why Our Dashboard Assistant Moved Off the Claude Code SDK

Platform

-- Prompt Management

-- Evaluations

-- Observability

-- Dataset Management

-- Prompt Chaining

Docs

Blog

Case Studies

Careers

Log in

Contact Us

Contact Us

Log In

Back

From Skills Back to Tools: Why Our Dashboard Assistant Moved Off the Claude Code SDK

By<br>Vladislav Vasilenko

May 20, 2026

illustration by Evan Casas

We just ripped out the Claude Code SDK and skills architecture powering Wrangler AI, our in-app assistant, and replaced it with a simple prompt and explicit tools. Here's why we made the switch.<br>Claude Code SDK powers Claude Code, the coding assistant that revolutionized how software engineers work. Wrangler AI isn't a coding agent. It's an assistant that helps users create prompts, evaluations, datasets, workflows, snippets, and so on, directly on our dashboard. Different surfaces, different physics.<br>What "powerful" actually cost us on the dashboard<br>The Claude Code SDK is great at what it's designed for: autonomous code agents, multi-step internal automation, large self-directed tasks. We wanted to see if we could lift that magic into a chat-style dashboard helper. We could — but at a price.<br>The cleanest way to see what changed is to ask the same user question in both versions and look at the traces side by side.<br>Question 1: "What is Docs Links?"<br>Claude Code SDK on the left, Tools on the rightUser asked the same question in both versions: a one-line lookup of a snippet by name.

"Docs Links / what's that?"<br>v131 — skills<br>v132+ — tools

Wrangler-turn wall-clock<br>3 min 4 sec<br>24 sec

LLM calls<br>20

Tool executions<br>19

Throwaway scripts written + executed

Mid-turn errors recovered from

Same question, same answer expected. One architecture took ~8× longer and made ~7× more tool/model calls.<br>Claude Code SDK<br>A wrangler-turn span wraps the entire user-visible turn. Inside it, the SDK opened a Claude Code session that ran:<br>"What is Docs Links?" using Claude Code SDKIt performed a long tool chain: ToolSearch to discover MCP tools, loaded a ~30KB search skill doc into context, then wrote and ran six different throwaway Python scripts (to make all of the api calls) to /tmp, recovering from a TypeError, a 405 on a folder endpoint, and two file-too-large errors trying to read its own 86KB persisted output back into context — before finally returning.<br>End-to-end wall-clock on that one wrangler-turn: 3 minutes 4 seconds to answer "what's that?". The model spent most of it deciding which tools existed, reading its own skill docs, writing throwaway Python to /tmp, and recovering from its own errors.<br>The Prompt with Tools<br>"What is Docs Links?" using ToolsIt only needed a simple flow: call search_entities to find the Docs Links snippet, call get_entity_details with the resolved snippet ID, and then return a short answer to the user.<br>End-to-end wall-clock: 24 seconds .<br>The traces show that v131 was making ~10× more LLM calls and taking ~8× longer per turn . Multiply that pattern across thousands of production requests and the aggregate cost and latency numbers stop being surprising — they're actually conservative.

Question 2: "Create me a prompt about weather"<br>Claude Code SDK on the left, Tools on the rightThis is the kind of "make me a thing" request the dashboard helper gets all day.

"Create me a prompt about weather"<br>v131 — skills<br>v132+ — tools

Wrangler-turn wall-clock<br>1 min 53 sec<br>38.5 sec

Nested Claude Code sessions

LLM calls<br>10

Tool executions

3× faster, end-to-end, on the same request. And the entire path is auditable from the trace — no skills loaded into context, no throwaway Python written to /tmp, no agent loop the user can't see.<br>The weather prompt request showed the same pattern on a creation task.<br>Claude Code SDK<br>"Create me a weather prompt" using Claude Code SDKThe skills version again ran through a nested Claude Code session. It took 1 minute 53 seconds , made 10 LLM calls , and executed 6 tools before the prompt was created.<br>Prompt with Tools<br>"Create me a weather prompt" using ToolsThe tools version used the direct product path: select the model config, create the input variable set, create the prompt, and create the prompt version.<br>End-to-end wall-clock: 38.5 seconds . It made 4 LLM calls and executed 4 named tools .<br>So this was not just a lookup problem. Across both a simple retrieval request and a common "make me a thing" request, the tools architecture gave us the same outcome with fewer side trips, fewer hidden loops, and much faster turns.<br>Why a prompt with explicit tools wins for dashboard work<br>No skill doc injected into context, no ToolSearch to discover MCP tools, no throwaway Python written to /tmp, no Bash running scripts that need recovery, no nested Claude Code session running its own loop on claude-sonnet-4-6. The model has its tools, knows their JSON schemas, and uses them — and every call shows up in the trace by name. Extremely simple.<br>That maps directly to what...