# AI use cases for Squadra (dagoSte/squadra) Repo: [dagoSte/squadra](https://github.com/dagoSte/squadra) | Use cases: 4 | Page: https://dedupedai.com/use-cases/dagoste-squadra/ Squadra is a free, local, MIT-licensed team toolkit that coordinates Claude Code, Codex and Cursor windows with shared tasks, proofs, reviews and a local dashboard. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Coordinate Claude, Codex and Cursor windows on one project: https://dedupedai.com/use-cases/dagoste-squadra/coordinate-ai-coding-windows-on-one-project/ - Pick the next window for a task by token use and quota: https://dedupedai.com/use-cases/dagoste-squadra/pick-the-next-ai-window-by-token-use/ - Prove each change with parallel tests and a source-bound proof: https://dedupedai.com/use-cases/dagoste-squadra/prove-changes-with-parallel-tests/ - Resume a long task in a fresh window without a full log dump: https://dedupedai.com/use-cases/dagoste-squadra/resume-long-task-in-fresh-window/ ## How to coordinate Claude, Codex and Cursor windows on one project with Squadra Repo: [dagoSte/squadra](https://github.com/dagoSte/squadra) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Give several AI coding windows shared tasks, acknowledgements and a review step on one codebase. **Who it is for:** A developer running two or more Claude Code, Codex or Cursor windows on the same project who keeps losing track of which window did what. **The problem:** Separate AI windows overlap, miss each other's changes and leave no clear record of who finished which task. **The outcome:** Each window works from a shared task list, confirms what it has picked up, and the dashboard shows progress, decisions and evidence in one place. ### Steps 1. Paste this into Claude Code, Codex or Cursor: Install Squadra by following https://github.com/dagoSte/squadra/blob/main/SETUP.md: show me the plan first, wait for my yes, then install it and check that it works. 2. Approve the plan the agent shows, then run `/reload-plugins` in Claude Code or start a new session in Codex or Cursor. 3. Run `squadra start` in the project folder to open the squad. (inferred, check the README) 4. Create work with `squadra task` and check the short summary with `squadra brief`. (inferred, check the README) 5. Follow incoming messages in each Claude window with `squadra inbox --follow`, and check open waits with `squadra status --all`. 6. Open the bilingual local dashboard to review progress, decisions and evidence. ### Prompt to paste into your agent ```text Using the Squadra repo at https://github.com/dagoSte/squadra, set up a shared squad for the project at with Claude Code and Codex windows. Show the install plan first and wait for my approval. Expected output: a list of open tasks and any waits for each window. Verify that `squadra status --all` runs without errors and lists every registered window. ``` ### Caveat Squadra needs Python 3.12 or newer and runs on macOS or POSIX only, as the README states. 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/dagoste-squadra/coordinate-ai-coding-windows-on-one-project/ ## How to pick the next window for a task by token use and quota with Squadra Repo: [dagoSte/squadra](https://github.com/dagoSte/squadra) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README See what each AI window spent and which window or model should take the next task. **Who it is for:** A developer whose squad mixes Claude Code and Codex windows and who wants to stop guessing which window has room for the next piece of work. **The problem:** Each provider shows usage only for its own tool, so there is no single view of spend, quota and load across the squad. **The outcome:** A ranked view of windows and models for a task, with the figures behind each suggestion. The tool assigns nothing; you decide. ### Steps 1. Install Squadra with the one-sentence setup from SETUP.md, as shown in the README. 2. Register your windows as the guide describes, so their sessions are counted with `--session`. 3. Run `squadra balance --task` to get the full ranking for a task. 4. Read the Usage page in the local dashboard to compare token totals per window and model. 5. Use `task`, `brief` or `start` for a short forecast and balance before you assign work. ### Prompt to paste into your agent ```text Using the Squadra repo at https://github.com/dagoSte/squadra, rank the AI windows in the squad at for task . Expected output: the output of `squadra balance --task` with the figures each suggestion rests on. Verify that the ranking lists every registered window and that no window is assigned automatically. ``` ### Caveat Usage figures are computed on your computer from the session files the clients keep, and the README says Balance assigns nothing. 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/dagoste-squadra/pick-the-next-ai-window-by-token-use/ ## How to prove each change with parallel tests and a source-bound proof with Squadra Repo: [dagoSte/squadra](https://github.com/dagoSte/squadra) | Difficulty: Advanced | Time: 45 minutes | Basis: From the README Run the test gate for a task, queue busy resources and attach proof tied to the frozen code. **Who it is for:** A developer or reviewer who needs evidence that a specific code state passed tests before a release or merge. **The problem:** Several windows running tests at once fight over the same gate, and a test result is hard to tie to the exact code it checked. **The outcome:** Each task gets a proof bound to its source, and windows wait in line for the shared test gate in the order they asked. ### Steps 1. Wait your turn for a busy shared resource with `squadra resource acquire --wait SECONDS`. (inferred, check the README) 2. Run the proof for the task with `squadra prove`, and check its options with `squadra prove --help`. 3. If one run should serve several tasks on the same frozen code, use `prove --also` only when every reader's Squadra lists it in `squadra prove --help`. 4. Confirm the proof and any final reads in the dashboard evidence view. 5. Run `squadra doctor` if the registry looks damaged; use `squadra doctor --quick` for a fast check. (inferred, check the README) ### Prompt to paste into your agent ```text Using the Squadra repo at https://github.com/dagoSte/squadra, prove task in the project at . Expected output: a proof record bound to the current source state. Verify with `squadra doctor --quick` that the registry is healthy before you rely on the proof. ``` ### Caveat The README says `prove --also` should only be used when every reader's Squadra supports it, so check the help output first. 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/dagoste-squadra/prove-changes-with-parallel-tests/ ## How to resume a long task in a fresh window without a full log dump with Squadra Repo: [dagoSte/squadra](https://github.com/dagoSte/squadra) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README Prepare a bounded resume packet so an agent can continue a long task from its references. **Who it is for:** A developer whose AI window has grown too long to continue a task well and who wants to hand the work to a new window. **The problem:** Pasting whole conversations or logs into a new window wastes context and makes the agent slower to resume. **The outcome:** A short, read-only packet with the task, its scope and proof references, so the new window picks up where the last one stopped. ### Steps 1. Run `squadra context --task ID` with the ID of the task to resume. 2. Open a new Claude Code, Codex or Cursor window in the same project. 3. Give the new window the packet and point it to the task references rather than the old conversation. 4. Check the task status with `squadra status --all` before the new window starts editing. 5. Use the native compaction in Codex or Claude when the window needs to shrink, as the context workflow describes. ### Prompt to paste into your agent ```text Using the Squadra repo at https://github.com/dagoSte/squadra, prepare a resume packet for task in the project at . Expected output: a read-only packet listing the task, scope and proof references. Verify that the packet contains references and not full conversation or log text. ``` ### Caveat The README says the packet neither clears provider cache nor guarantees token savings, so measure your own usage. 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/dagoste-squadra/resume-long-task-in-fresh-window/