# AI use cases for Deer Workflow (deerwork-ai/deer-workflow) Repo: [deerwork-ai/deer-workflow](https://github.com/deerwork-ai/deer-workflow) | Use cases: 5 | Page: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/ deer-workflow is an open-source TypeScript runtime for building observable, reusable Agent graphs, where code defines the steps and Coding Agents do the work inside each step. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Generate a runnable TypeScript workflow from plain English: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/generate-typescript-workflow-from-description/ - Run a workflow in CI and parse its JSON event stream: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/run-workflow-in-ci-with-json-events/ - Run the Deep Research example to produce an HTML report: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/try-deep-research-example/ - Generate a workflow with Claude Code or Pi instead of Codex: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/swap-agent-runtime-for-generation/ - Add an integration for a new coding agent: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/add-new-agent-integration/ ## How to generate a runnable TypeScript workflow from plain English with Deer Workflow Repo: [deerwork-ai/deer-workflow](https://github.com/deerwork-ai/deer-workflow) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Describe a multi-step job in one sentence and get a runnable TypeScript workflow module. **Who it is for:** A developer or analyst who wants a multi-step AI pipeline, such as parallel research plus a written report, but does not want to hand-write the orchestration code. **The problem:** Multi-step agent jobs are hard to repeat and review when the control flow lives inside a long chat with an AI agent. **The outcome:** A workflow.ts file that holds the execution path in readable TypeScript, which you can run, inspect, and edit. ### Steps 1. Install the released CLI with `bun install --global @deerwork-ai/deer-workflow`. 2. Make sure you are signed in to Codex CLI, which is the default Agent runtime. 3. Generate the module with `deer-workflow create "Create a Workflow that accepts a topics string array, researches each topic in parallel, and synthesizes a report" > workflow.ts`. (inferred, check the README) 4. Run it with its example input: `deer-workflow run ./workflow.ts --input '{"topics":["Agent Skills","Dynamic Workflows"]}'`. (inferred, check the README) 5. Open workflow.ts and review the phases, inputs, and failure handling before you change anything. ### Prompt to paste into your agent ```text Using the deer-workflow repository (deerwork-ai/deer-workflow), generate a runnable TypeScript Workflow for with the deer-workflow create command, save it to workflow.ts, and run it with a sample input. Expected output: a workflow.ts file plus a run result. Verify that the run finishes and that the phases in the file match the steps described in the job. ``` ### Caveat The README says Codex is the default runtime and Claude Code and Pi are built in, so the generation step depends on an installed and signed-in Agent CLI. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/generate-typescript-workflow-from-description/ ## How to run a workflow in CI and parse its JSON event stream with Deer Workflow Repo: [deerwork-ai/deer-workflow](https://github.com/deerwork-ai/deer-workflow) | Difficulty: Advanced | Time: 45 minutes | Basis: From the README Execute a saved workflow in a build pipeline and read a stable JSONL event stream. **Who it is for:** A DevOps or platform engineer who needs an agent-driven workflow to run on a server or in a CI job without an interactive terminal. **The problem:** Agent runs in a chat window cannot be scheduled, logged, or checked by automation. **The outcome:** A non-interactive run that prints one JSON event per line, which a pipeline can capture, store, and act on. ### Steps 1. Install Bun and the released CLI with `bun install --global @deerwork-ai/deer-workflow`. 2. Commit your reviewed workflow.ts file to the repository that your pipeline uses. 3. Add a pipeline step that runs `deer-workflow run ./workflow.ts --input '' --print`. (inferred, check the README) 4. Capture stdout and parse each line as a JSON event. 5. Fail the job or alert the team based on the events that the run reports. ### Prompt to paste into your agent ```text Using the deer-workflow repository (deerwork-ai/deer-workflow), add a CI step that runs ./workflow.ts with the input and the --print flag. Expected output: a captured JSONL event stream saved as an artifact. Verify that every stdout line parses as JSON before the job reports success. ``` ### Caveat The README describes the JSONL stream as stable, but it does not show the event schema, so check the API Reference before writing parsers against specific event names. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/run-workflow-in-ci-with-json-events/ ## How to run the Deep Research example to produce an HTML report with Deer Workflow Repo: [deerwork-ai/deer-workflow](https://github.com/deerwork-ai/deer-workflow) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Clone the repo and run a bundled example that researches angles in parallel and builds a report. **Who it is for:** A researcher or analyst who wants to see a finished multi-agent research workflow before writing their own. **The problem:** It is hard to judge what a workflow runtime can do without seeing a complete example that runs end to end. **The outcome:** An interactive HTML report produced by a workflow that discovers angles, investigates them in parallel, and verifies claims. ### Steps 1. Clone the repository with `git clone https://github.com/deerwork-ai/deer-workflow.git`. 2. Change into the folder with `cd deer-workflow` and install dependencies with `bun install`. 3. Open examples/deep-research/README.md and follow its documented commands. 4. Watch the phases in the live TUI while the example runs. 5. Open the generated HTML report and check its claims against the sources the run used. ### Prompt to paste into your agent ```text Using the deer-workflow repository (deerwork-ai/deer-workflow), follow examples/deep-research/README.md to run the Deep Research example on the topic . Expected output: an interactive HTML report. Verify that the report includes the claims verification step described in the example README. ``` ### Caveat The README says the examples need the repository cloned or downloaded first, so the example commands are not available from the npm package alone. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/try-deep-research-example/ ## How to generate a workflow with Claude Code or Pi instead of Codex with Deer Workflow Repo: [deerwork-ai/deer-workflow](https://github.com/deerwork-ai/deer-workflow) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Switch the Agent runtime used to generate a workflow without changing the rest of the process. **Who it is for:** A developer on a team that already uses Claude Code or Pi and does not want to add Codex just to generate workflows. **The problem:** A tool tied to one Agent vendor forces extra setup for teams that use a different coding agent. **The outcome:** A workflow.ts file generated by the Agent runtime the team already has installed. ### Steps 1. Install the released CLI with `bun install --global @deerwork-ai/deer-workflow`. 2. Make sure the Agent CLI you want to use is installed and signed in. 3. Generate the module with `deer-workflow create "" --agent claude > workflow.ts`, or use `--agent pi` for Pi. (inferred, check the README) 4. Run the file with `deer-workflow run ./workflow.ts --input ''`. (inferred, check the README) 5. Compare the generated phases with your goal and edit workflow.ts if needed. ### Prompt to paste into your agent ```text Using the deer-workflow repository (deerwork-ai/deer-workflow), generate a Workflow for with the --agent claude option and save it to workflow.ts. Expected output: a runnable TypeScript module. Verify that the file runs with a sample input and that the phases match the job. ``` ### Caveat The README names Claude Code and Pi as built-in Harnesses, but it does not say how well each one performs on the same workflow description, so test the output before you rely on it. Risk, credentials: Generation uses the Agent CLI account you sign in with, and the README does not describe how those credentials are stored or shared. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/swap-agent-runtime-for-generation/ ## How to add an integration for a new coding agent with Deer Workflow Repo: [deerwork-ai/deer-workflow](https://github.com/deerwork-ai/deer-workflow) | Difficulty: Advanced | Time: 90 minutes | Basis: From the README Clone the repo, run the checks, and build a new Agent integration beside the built-in ones. **Who it is for:** An open-source contributor or platform engineer who wants Deer Workflow to work with a coding agent the project does not support yet. **The problem:** A workflow runtime is only useful for teams whose preferred coding agent is supported, so new integrations are needed. **The outcome:** A tested pull request that adds an Agent integration beside ClaudeAgent and PiAgent. ### Steps 1. Clone the repository with `git clone https://github.com/deerwork-ai/deer-workflow.git`. 2. Install dependencies and Git hooks with `bun install`. 3. Run the CLI from source with `bun run dev -- --help` to confirm the setup works. 4. Read the public Agent interface and the existing ClaudeAgent and PiAgent integrations before you write new code. 5. Run the full quality gate with `bun run check` before you open a pull request. ### Prompt to paste into your agent ```text Using the deer-workflow repository (deerwork-ai/deer-workflow), read the public Agent interface and the ClaudeAgent and PiAgent integrations, then draft a new Agent integration for . Expected output: a pull request branch with the new integration. Verify that bun run check passes before you report the work as done. ``` ### Caveat The README invites integrations for other Coding Agents but does not describe the Agent interface itself, so read the source and the API Reference before designing the new integration. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/deerwork-ai-deer-workflow/add-new-agent-integration/