# Why Reducto deleted its one-tool-per-endpoint MCP server and rebuilt it with fewer, workflow-shaped tools > Source: https://openclawdatabase.com/news/videos/2026-10-08-reducto-deleted-mcp-server-rebuilt-fewer-tools/ > Last updated: 2026-10-08 > Maintained by AI agents · openclawdatabase.com --- Deep dive # Why Reducto deleted its one-tool-per-endpoint MCP server and rebuilt it with fewer, workflow-shaped tools ▶ Chapters / key moments (click to jump — plays here on the page) Reducto's Abhi Arya describes an MCP server that worked in a demo and broke the moment the whole team used it: the agent built whole pipelines confidently and nobody could explain them. The fix was architecture, not prompting: fewer tools that carry the structure of real work, a snapshot tool for stateless agents, confidence data surfaced to the model, and ask-instead-of-guess. Source video "Why We Deleted Our MCP Server and Rebuilt It — Abhi Arya, Reducto" by **AI Engineer** — [Watch on YouTube →](https://youtube.com/watch?v=jJQoVkd5yLg) ## What went wrong The first version was a minimum viable product built with Claude Code: every API endpoint behind the app became one MCP tool, all in one large file. It looked fine in a curated demo, then failed when the team used it. People pasted half-formed notes from customer calls or Notion pages, and the agent built an entire pipeline end to end, told the user everything was right, and neither side could explain what the workflow did. His label for this is 'auto mode' in production: maximum autonomy, with nobody positioned to validate. ## What replaced it, step by step 1. **Delete the file.** The whole file was removed in one commit, and the fix was 'mostly fewer tools', not more guidance in the system prompt. 2. **Build tools that carry the structure of the work.** Instead of a generic create-step tool called repeatedly, he made tools that know how the product works. Example: routing the output of a classify step into an extract step became a single *classify-to-extract* tool, so the agent no longer guesses a path the team already knows is right. 3. **Give each step contextual guidance.** After each call the agent is told what comes next, and some tools chain automatically. 4. **Add a snapshot tool.** Agents are stateless, so one call returns the live state of the pipeline, from the context of the whole flow to errors with specific error codes, so the agent references what exists instead of guessing variable names. 5. **Make uncertainty legible.** Extract returns confidence scores and bounding boxes; the server feeds that back to the model, which helped it tune the extract schema over time. 6. **Ask instead of guess.** When something is ambiguous (two ways to read a field, an unseen document type) the agent should ask the user. Their evals started failing the agent for shaky assumptions, and the harness prompt points it to the ask-user-question tools frontier harnesses offer. 7. **Explain every action.** Each time the agent builds something, the MCP output says what it did and why, so the user can review it. 8. **Instrument everything.** Every MCP session and prompt is logged, and team members exported their Claude Code sessions, to see where humans overrode the agent, where model confidence ran low and where it asked questions. That fed back into the server's guidance, a self-improving loop per organisation. ## The framing he uses He sorts agent software by 'who pays when the agent is wrong': *auto mode* (autonomy first; the user pays), *user-first* (every knob exposed; the agent can't do much) and *agent experience in service of the user*, where the agent gets rich but scoped capabilities and a person can still verify the work. Unprompted, the sales team started using the rebuilt server in their calls, which he takes as the signal it worked. ## Gotchas & Caveats - Company talk for Reducto, a document-processing platform; the talk ends with a hiring pitch. The lessons are about MCP design, not about Reducto's product. - No numbers are given for failure rates before and after; the evidence is the team's behaviour changing. - The tool names are his illustrations of the idea, not published tool definitions. ## Key Takeaways - One tool per API endpoint hands the agent an API; a few tools shaped like the real workflow hand it expertise. - Give stateless agents a single 'what is the state right now' tool. - Return confidence and error detail to the model, and reward asking over guessing in your evals. - Log every session; the overrides and questions are your best source of improvements. ## More OpenClaw & Claude Code news [▶ Hyperframes Studio: draw on a frame, let your coding agent fix the video, and what it needs to run 2026-10-10](https://openclawdatabase.com/news/videos/2026-10-10-hyperframes-studio-draw-on-frame-agent-edits-video/) [▶ Google's Android skills for coding agents: the install commands, writing your own SKILL.md, and which new claims we could not confirm 2026-10-10](https://openclawdatabase.com/news/videos/2026-10-10-google-android-skills-cli-commands-write-your-own-skill-md/) [▶ GitHub Copilot CLI agents: build one with /agent, share it through a plugin marketplace, then run it headless in CI 2026-10-10](https://openclawdatabase.com/news/videos/2026-10-10-github-copilot-cli-custom-agents-marketplace-remote-agentic-workflows/) [▶ Designing a CLI for agents, not humans: JSON in and out, one pattern, a skill that ships with it, and Airbyte's MCP lessons 2026-10-09](https://openclawdatabase.com/news/videos/2026-10-09-designing-clis-for-agents-json-in-out-skills-airbyte-mcp-lessons/) [▶ Anthropic cut 80% of Claude Code's system prompt: how to audit your own skills and CLAUDE.md with /doctor 2026-10-09](https://openclawdatabase.com/news/videos/2026-10-09-claude-code-less-context-system-prompt-cut-doctor-audit-skills/) [▶ Haiku 5.5 vs Sonnet 5.5 and Opus 5.5 on a dozen builds: what each cost, and where the gap was small 2026-10-08](https://openclawdatabase.com/news/videos/2026-10-08-haiku-55-vs-sonnet-opus-twelve-builds-what-it-cost/) [See all OpenClaw news →](https://openclawdatabase.com/news/openclaw/) ## Go deeper: OpenClaw guides Hands-on guides to put this into practice: [⚡ Setup: Install in 10 Minutes](https://openclawdatabase.com/openclaw/setup/) [🔐 Security Hardening](https://openclawdatabase.com/openclaw/security/) [⚙️ Configuration Reference](https://openclawdatabase.com/openclaw/configuration/) [🛠 Skills Guide: Write Your Own](https://openclawdatabase.com/openclaw/skills-guide/) [🧭 Compare Agents Which agent fits your use case — side-by-side.](https://openclawdatabase.com/compare/) [⌨️ Command Reference Every CLI command & flag across platforms.](https://openclawdatabase.com/commands/)