Source: https://internal-agents.com/agents/amplitude-design-agent # Amplitude — Design Agent Amplitude built Design Agent, an internal web tool that anyone in the company reaches behind Google OAuth to turn a text prompt, a screenshot, or a rough product idea into on-brand HTML output. Amplitude's brand guidelines and design-system tokens sit in the agent system prompt, so the output reads as the company's own. The tool began as a proof of concept for product and engineering teams and has expanded to other designers, marketers, and creative teams. - Company: [Amplitude](https://internal-agents.com/organizations/amplitude) - Collection: Agents - Approach type: Agent - Deployment stage: Deployed - Autonomy: Drafts reviewed - Evidence strength: Limited primary - Status: Internal - First reported year: 2026 - Work: Design - Interfaces: Web - Invocation: Interactive - Entry reviewed: 2026-09-21 ## Purpose ### Summary Amplitude built Design Agent, an internal web tool that anyone in the company reaches behind Google OAuth to turn a text prompt, a screenshot, or a rough product idea into on-brand HTML output. Amplitude's brand guidelines and design-system tokens sit in the agent system prompt, so the output reads as the company's own. The tool began as a proof of concept for product and engineering teams and has expanded to other designers, marketers, and creative teams. Fact · Reported · High confidence · `amplitude-design-agent--summary` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 18, 20, 54, 64 Representative workflow: Prompt or screenshot in the internal web app, through managed-agent generation and R2 storage, to an artifact a person shares or carries forward. ## How it works ### Submit the request in the web interface The interface offers a text input, an option to upload a screenshot or reference image, and a panel that displays the generated artifact. Fact · Reported · High confidence · `amplitude-design-agent--primitives-1` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, line 54 ### Let the managed agent plan and generate Claude Managed Agents reasons about the prompt, runs a multi-step plan, handles tool-call follow-ups, and assembles the final artifact. Fact · Reported · High confidence · `amplitude-design-agent--primitives-2` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 28, 81 ### Persist the artifact with a stable URL Every generated artifact lands in Cloudflare R2 with a stable, permanent URL, and users keep a visible generation history. Fact · Reported · High confidence · `amplitude-design-agent--primitives-3` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 30, 56, 85 ### Share, iterate, or carry the output forward People share the output links in Slack threads, return to generate follow-ups on previous outputs, or copy the generated HTML into a coding agent. Fact · Reported · High confidence · `amplitude-design-agent--primitives-4` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 56, 71, 75 ## Where people stay involved - **prompt or screenshot → generated HTML artifact a person judges, shares, or carries forward** — Work-product review · Level 3 **Reported:** A person writes every request and then judges the artifact, sharing the link, re-prompting, or pasting the HTML into a coding agent; the post documents no separate approval gate. ### Supervision evidence #### Operating model assessment Level 3 for prompt or screenshot → generated HTML artifact a person judges, shares, or carries forward; human attention boundary: work-product-review. Inference · Catalog judgment · Medium confidence · `amplitude-design-agent--operating-models-0` Confidence reason: The source shows the agent producing a finished artifact that the requester then judges, shares in Slack, re-prompts, or copies into a coding agent; it documents no separate approval gate, so the boundary is read from the described use rather than from a stated review step. Qualifications: - Observation date: 2026-05 Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 71, 73, 75 ## Implementation details ### Harness Claude Managed Agents supplies reasoning, tool use, and multi-step generation; Amplitude wrote no state machine, prompt chain, or tool-calling logic. A thin wrapper of Cloudflare Workers serves the web interface and the agent interaction endpoint. Fact · Reported · High confidence · `amplitude-design-agent--architecture-harness` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 28-29, 81, 85 ### Model Claude, through Claude Managed Agents; no model version is named. Fact · Reported · High confidence · `amplitude-design-agent--architecture-model` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 18, 22, 28 ### Interfaces web Fact · Reported · High confidence · `amplitude-design-agent--architecture-interfaces` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 54, 85 ### Tool access Amplitude states that it defined the agent's tools and context but does not name them. Generated artifacts are written to Cloudflare R2, each with a permanent URL. Fact · Reported · High confidence · `amplitude-design-agent--architecture-tool-access` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 30, 81, 85 ### Knowledge A structured Markdown document in the Google design.md format holds Amplitude's brand guidelines, color system, typography rules, spacing conventions, and component patterns. Fact · Reported · High confidence · `amplitude-design-agent--architecture-knowledge` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 46, 48 ### Context management The brand and design-system document is baked into the agent system prompt, so it is present in every conversation; behavior changes ship by editing that prompt and take effect in minutes. Fact · Reported · High confidence · `amplitude-design-agent--architecture-context-mgmt` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 48, 83 ### Prompt or screenshot to on-brand HTML artifact A text prompt, a screenshot, or a rough product idea becomes interactive HTML styled to the Amplitude design system. Fact · Reported · High confidence · `amplitude-design-agent--primitives-0` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 18, 20 ### Bake brand context into the system prompt Amplitude's design philosophy and design-system tokens are written into the system prompt, which Amplitude reports turned generic model output into output that reads as its own. Fact · Reported · High confidence · `amplitude-design-agent--primitives-5` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 18, 46, 48, 50 ### Implementation coverage - **Model:** reported - **Harness:** reported - **Sandbox:** unreported — The post names Claude Managed Agents as the runtime but no execution isolation boundary; the legacy unknown claim stays in the research details. - **Tool access:** reported - **Knowledge:** reported - **Context management:** reported - **Credentials:** unreported — Google OAuth gates human access to the web app; how the agent itself authenticates to Claude Managed Agents, Workers, or R2 is not documented. - **Interfaces:** reported ## Validation and failure handling ### Instrument every session and turn failures into regression tests Amplitude tracks every session, output, and tool call with its own analytics, reviews sessions to find failure patterns, and creates eval examples from the worst outputs as regression tests before updating skills files, prompts, and agent configuration. Fact · Reported · High confidence · `amplitude-design-agent--primitives-6` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 91, 93 ## Reported observations Observation: Adoption output · Reported measurement · Design Agent session snapshots across Amplitude teams in the first few weeks of use ### Headline claim Over 2,219 session snapshots in Design Agent's first few weeks, with repeat usage across multiple teams (self-reported) Metric · Reported · Medium confidence · `amplitude-design-agent--headline-metric` Confidence reason: Amplitude reports the count from its own analytics instrumentation on Design Agent, with no independent review and no total of eligible users, teams, or sessions. Qualifications: - Reported by: Amplitude - Scope: Design Agent session snapshots across Amplitude teams in the first few weeks of use, as reported in the May 2026 post - Denominator: None reported; the post gives no user, team, or session total to divide by - Method: Amplitude product analytics instrumented on Design Agent, which the post says tracks every session, output, and tool call - Observation date: 2026-05 Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, line 64 Observation: Adoption output · Estimate · Ratio of viewers of generated artifacts to the people who generated them ### Key observation Amplitude reports roughly 2-4x more viewers than makers for generated artifacts, which it reads as a signal that outputs are shared beyond the people who generate them Metric · Reported · Low confidence · `amplitude-design-agent--key-metrics-0` Confidence reason: The post gives the ratio as roughly 2-4x with no period, no counting rule for a viewer or a maker, and no underlying totals, so the range is an approximation rather than a stated measurement. Qualifications: - Reported by: Amplitude - Scope: Viewers of generated Design Agent artifacts compared with the people who generated them - Denominator: None reported - Method: Not stated; the post reports the ratio without a counting rule - Observation date: 2026-05 Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, line 75 Observation: Implementation scale · Qualitative · Elapsed build time to a working Design Agent for product and engineering teams ### Key observation Amplitude reports a working Design Agent for its product and engineering teams two days after starting, on Claude Managed Agents and Cloudflare Metric · Reported · Medium confidence · `amplitude-design-agent--key-metrics-1` Confidence reason: The author states the two-day figure twice as elapsed build time for the first working version; it describes implementation effort rather than a performance result, and no start date or effort total is given. Qualifications: - Reported by: Amplitude - Scope: Elapsed time from starting the build to a working Design Agent for Amplitude product and engineering teams - Denominator: Not applicable - Method: Not stated; the author reports the duration without a record of hours or a defined start - Observation date: 2026-05 Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 22, 107 ## Lessons ### Lesson Amplitude wrote its brand guidelines and design-system tokens into the agent system prompt and reports that the difference from generic model output was drastic; this is its own result for one internal design tool, not a measured comparison. Fact · Reported · Medium confidence · `amplitude-design-agent--lessons-learned-0` Confidence reason: The post describes the design.md context document and the system-prompt change as steps the team took, and calls the difference drastic; that judgment is the author's own, with no side-by-side evaluation. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 46, 48, 50 ### Lesson Amplitude instruments Design Agent with its own product analytics, reviews sessions to find failure patterns, and turns the worst outputs into eval examples that act as regression tests. Fact · Reported · High confidence · `amplitude-design-agent--lessons-learned-1` Confidence reason: The measurement section states the practice directly: sessions reviewed in Agent Analytics for failure patterns, eval examples created from the worst outputs as regression tests, then skills files, prompts, and configuration updated. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 91, 93 ### Lesson Amplitude reports that managed agent orchestration and serverless hosting left it with no servers, monitoring, autoscaling, or database to run, which suits an internal tool maintained by one person. Fact · Reported · Medium confidence · `amplitude-design-agent--lessons-learned-2` Confidence reason: The hosting section states the absence of servers, monitoring, autoscaling, and a database and ties it to a one-person internal tool; this is the team's own description of its setup, not an audited operating record. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, line 87 ### Lesson Will Newton concludes that agents are products needing repeated iteration with real users, and advises keeping the stack simple and not skipping instrumentation; this is his stated view from building this one internal tool. Opinion · Reported · Medium confidence · `amplitude-design-agent--lessons-learned-3` Confidence reason: The closing section is the author's stated conclusion from this one build; he offers no comparison with other agents, so the advice is attributed rather than established. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 101-110 ## Reviewed legacy details ### Sandbox unknown Inference · Catalog judgment · Medium confidence · `amplitude-design-agent--architecture-sandbox` Confidence reason: The post says Claude Managed Agents handles execution behind the scenes but names no execution isolation boundary for the agent's work; unknown records the gap, not an absent sandbox. Evidence: - Supports · [1] [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/amplitude-design-agent-source-1/content.md) · Preserved content.md, lines 28, 81 ## Question coverage and scope - **purpose:** Reported - **workflow:** Reported - **human involvement:** Reported — A person writes every request and then judges the artifact, sharing the link, re-prompting, or pasting the HTML into a coding agent; the post documents no separate approval gate. - **implementation:** Reported - **validation:** Reported - **observations:** Reported - **lessons:** Reported ## Sources 1. [How we built a design agent at Amplitude with Claude managed agents and Cloudflare](https://www.amplitude.com/blog/design-agent) - Engineering blog · First party · Evidence - Original URL: - Publisher: Amplitude · Published: 2026-05-19 · Accessed: 2026-09-21 · Last verified: 2026-09-21 - Preserved copy in the repository: