# AI use cases for Strands Decider (strands-labs/strands-decider) Repo: [strands-labs/strands-decider](https://github.com/strands-labs/strands-decider) | Use cases: 4 | Page: https://dedupedai.com/use-cases/strands-labs-strands-decider/ Strands Decider is a small, fast decision model that picks between options, answers yes or no, or rates text on a scale, with a confidence score on each answer. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Route support messages to the right team: https://dedupedai.com/use-cases/strands-labs-strands-decider/route-tickets-to-a-team/ - Flag messages that convey urgency: https://dedupedai.com/use-cases/strands-labs-strands-decider/flag-urgent-messages-yes-no/ - Serve a scorer that other services can call: https://dedupedai.com/use-cases/strands-labs-strands-decider/score-tone-over-http/ - Check a form image for a filled-in field: https://dedupedai.com/use-cases/strands-labs-strands-decider/check-form-images/ ## How to route support messages to the right team with Strands Decider Repo: [strands-labs/strands-decider](https://github.com/strands-labs/strands-decider) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Pick the owning team for an incoming message, with a confidence score on the choice. **Who it is for:** A support or operations lead, or an agent that triages inbound messages, when each message needs an owner and nobody has time to read every one. **The problem:** Manual triage is slow, and a keyword rule misses messages that describe a problem without naming the team. **The outcome:** Each message gets one of the listed teams plus a confidence value, so confident routes can go straight through and low-confidence ones go to a person. ### Steps 1. `pip install strands-decider` 2. `strands-decider ask StrandsAgents/strands-decider-2B-qwen3.5-v1-2610 --state "Help! My payouts have been failing for 3 days! " --choice "Which team should handle this?=billing,sales,retail"` (inferred, check the README) 3. Read the line that starts with choice_0 and the confidence in brackets. 4. Send answers at confidence 0.9 or more straight through. Send lower ones to a person to confirm. ### Prompt to paste into your agent ```text Using the strands-decider repo, run the ask command with --state set to and a --choice question listing the team names . Return the chosen team and its confidence. Verify that the confidence is a number between 0 and 1 and that the chosen team is one of the listed options. ``` ### Caveat The README says answers at confidence 0.9 or more are right about 95% of the time on short classification tasks it has never seen, so routing on lower confidence needs a human check. 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/strands-labs-strands-decider/route-tickets-to-a-team/ ## How to flag messages that convey urgency with Strands Decider Repo: [strands-labs/strands-decider](https://github.com/strands-labs/strands-decider) | Difficulty: Beginner | Time: 15 minutes | Basis: From the README Get a yes or no urgency check on each message from a single command. **Who it is for:** An agent or developer building a queue that should surface urgent messages first. **The problem:** Urgency is a judgement that depends on wording, not on a fixed keyword, so simple filters miss many urgent messages. **The outcome:** Each message gets a noul score between 0 and 1, where values closer to 1 lean more toward Yes. ### Steps 1. `pip install strands-decider` 2. `strands-decider ask StrandsAgents/strands-decider-2B-qwen3.5-v1-2610 --state "Help! My payouts have been failing for 3 days! " --noul "Does this convey urgency?"` (inferred, check the README) 3. Read the noul_0 value. Treat values near 1 as urgent and values near 0 as not urgent. 4. Put messages above your chosen cutoff at the top of the queue. ### Prompt to paste into your agent ```text Using the strands-decider repo, run the ask command with --state set to and --noul set to "Does this convey urgency?". Return the noul value. Verify that the value is between 0 and 1 and that a clearly urgent test message scores higher than a routine one. ``` ### Caveat The README gives no fixed cutoff for urgency, so you need to pick and test your own threshold on your own messages. 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/strands-labs-strands-decider/flag-urgent-messages-yes-no/ ## How to serve a scorer that other services can call with Strands Decider Repo: [strands-labs/strands-decider](https://github.com/strands-labs/strands-decider) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Run the decision model as a local HTTP server and send it questions with curl. **Who it is for:** A developer whose agent or other service needs to call a decision model over HTTP for tone or sentiment scores on each request. **The problem:** Every service that needs a score would otherwise load its own model or call a slower general LLM for a narrow question. **The outcome:** A local server on port 8000 returns JSON with a score and token usage for each question, ready for another service to use. ### Steps 1. `pip install strands-decider` 2. `strands-decider serve StrandsAgents/strands-decider-2B-qwen3.5-v1-2610 --port 8000` 3. `curl -s localhost:8000/v1/systemone -H 'content-type: application/json' -d '{"state": "Help! My payouts have been failing for 3 days!", "questions": {"is_urgent": {"type": "noul", "instructions": "Does this convey urgency?"}}}'` (inferred, check the README) 4. Read the answers object in the JSON response and parse the noul value for your caller. ### Prompt to paste into your agent ```text Using the strands-decider repo, start the server with the 2B model on port 8000, then POST a state of and a score question to /v1/systemone. Return the score and confidence from the JSON response. Verify that the response contains an answers object and a latency_ms field. ``` ### Caveat The README describes running the server locally, and says to see its deployment guide before letting other people or services call it from elsewhere. 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/strands-labs-strands-decider/score-tone-over-http/ ## How to check a form image for a filled-in field with Strands Decider Repo: [strands-labs/strands-decider](https://github.com/strands-labs/strands-decider) | Difficulty: Advanced | Time: 60 minutes | Basis: From the README Ask a yes or no question about an image, such as whether a signature block is filled in. **Who it is for:** An operations or document-review person who needs to confirm that scanned pages have a specific field completed. **The problem:** Checking many scanned pages by eye takes time, and a text-only model cannot see the page. **The outcome:** Each page gets a noul score for the question, so pages that fail the check can be sent back for review. ### Steps 1. `pip install "strands-decider[vision]"` 2. Install transformers 5.18 or later, as the README requires for vision. 3. `strands-decider ask StrandsAgents/strands-decider-2B-qwen3.5-v1-2610 --state "" --image page.png --noul "Is the signature block filled in?"` (inferred, check the README) 4. Review pages with low scores by hand. ### Prompt to paste into your agent ```text Using the strands-decider repo with the vision extra installed, run the ask command on with a noul question about . Return the noul value. Verify the result on two known pages, one filled in and one blank, before trusting the scores. ``` ### Caveat The README says the vision support works with the published checkpoints without image training, and that the Gemma 4 models read text only, so use the Qwen3.5 model for images. 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/strands-labs-strands-decider/check-form-images/