# AI use cases for Asamana (dreamscroll-lab/asamana) Repo: [dreamscroll-lab/asamana](https://github.com/dreamscroll-lab/asamana) | Use cases: 4 | Page: https://dedupedai.com/use-cases/dreamscroll-lab-asamana/ Asamana is an open-source Python and React engine that generates a themed world of AI characters that act on their own and produce a story without a script. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Replay a finished story without any API keys: https://dedupedai.com/use-cases/dreamscroll-lab-asamana/replay-example-story-without-keys/ - Generate and run a new themed AI story: https://dedupedai.com/use-cases/dreamscroll-lab-asamana/generate-and-run-themed-story/ - Audit a finished world's narrative quality: https://dedupedai.com/use-cases/dreamscroll-lab-asamana/audit-narrative-quality-of-a-world/ - Trace every LLM call to debug a run: https://dedupedai.com/use-cases/dreamscroll-lab-asamana/trace-llm-calls-to-debug-a-run/ ## How to replay a finished story without any API keys with Asamana Repo: [dreamscroll-lab/asamana](https://github.com/dreamscroll-lab/asamana) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Start Asamana locally and browse the two bundled example worlds with no keys. **Who it is for:** A curious reader or developer who wants to see how an emergent story looks before paying for any model calls. **The problem:** Setting up an AI simulation before seeing output is a costly guess, and most of the tool needs keys to be useful. **The outcome:** The user has the app running locally and has browsed the map, characters, relationships and narrative feed of a finished example world. ### Steps 1. Run `git clone https://github.com/dreamscroll-lab/asamana.git && cd asamana` 2. Run `./deploy/asamana.sh start` 3. When the setup wizard appears, skip the optional API key configuration 4. Open the app in your browser and select the Xuanwu Gate Incident example world 5. Use the map, character, relationship and narrative feed views to replay the story ### Prompt to paste into your agent ```text Clone the repo at https://github.com/dreamscroll-lab/asamana, then run ./deploy/asamana.sh start without configuring keys. Report the name of one example world you can browse and the number of characters it lists. Verify that the backend starts without an error about missing API keys. ``` ### Caveat Creating, running, stepping or resetting a world requires API keys, so this path only covers browsing and replay. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/dreamscroll-lab-asamana/replay-example-story-without-keys/ ## How to generate and run a new themed AI story with Asamana Repo: [dreamscroll-lab/asamana](https://github.com/dreamscroll-lab/asamana) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README Turn one theme into a cast of AI characters and watch their story unfold step by step. **Who it is for:** A writer or game designer who wants a fresh, unscripted storyline built from a single theme and has a model API key. **The problem:** Writing branching characters and plot by hand takes a long time, and scripted tools lock the plot in advance. **The outcome:** A new world is created from your theme, reviewed and confirmed, and run for a chosen number of steps, producing a narrative feed. ### Steps 1. Run `./deploy/asamana.sh start` and enter your model vendor's API key in the setup wizard, or run `./deploy/asamana.sh keys` later 2. Choose a model with a flash-tier or smaller model for the first run, following the Choosing a model section 3. Open the app, enter a theme, and review and confirm the generated world 4. On the run page, enter a small step count such as 3 and start the run 5. Read the narrative feed and check the token usage against your vendor console ### Prompt to paste into your agent ```text In the repo https://github.com/dreamscroll-lab/asamana, use a configured API key to start the app and create a world from the theme . Run 3 steps and report the number of characters and the first three narrative lines. Verify that the run stops after the step count you set and does not run indefinitely. ``` ### Caveat Each step makes several LLM calls per character, so a 35-step world can use roughly 2.5M to 4.2M tokens, so set a spending cap first. Risk, credentials: The API has no authentication, so anyone who can reach it can use your configured keys; keep the default localhost binding or add access controls before exposing it. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/dreamscroll-lab-asamana/generate-and-run-themed-story/ ## How to audit a finished world's narrative quality with Asamana Repo: [dreamscroll-lab/asamana](https://github.com/dreamscroll-lab/asamana) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Use the developer audit tools to have an LLM review a completed world's story and cognition. **Who it is for:** A developer or prompt tuner who has a completed world and wants a structured review of its narrative before changing prompts. **The problem:** Judging narrative quality by reading raw logs is slow, and tuning prompts without a baseline makes regressions hard to see. **The outcome:** The user has an audit report for the whole narrative, each character's cognition, or world initialization, based on the audit types the README names. ### Steps 1. Enable the developer tools as described in docs/development.md#developer-tools 2. Run `./deploy/asamana.sh start` with API keys configured 3. Open the `#/dev` page in the app 4. Open the Audit section and run one audit on a completed world using `mams` for the whole narrative, `sams` for each character's cognition, or `init` for world initialization 5. Read the audit report and note the issues to fix in your prompts ### Prompt to paste into your agent ```text In the repo https://github.com/dreamscroll-lab/asamana, enable the developer tools and run an audit of type on the world . Report the main issues found in the audit report. Verify that the audit ran against a completed world rather than a world still in progress. ``` ### Caveat The README says audit and prompt replay tools call a model, so they need keys and are meant for local development. Risk, credentials: Audit runs call your configured model, which uses your API keys and spends your token budget. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/dreamscroll-lab-asamana/audit-narrative-quality-of-a-world/ ## How to trace every LLM call to debug a run with Asamana Repo: [dreamscroll-lab/asamana](https://github.com/dreamscroll-lab/asamana) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Inspect prompts, responses, tokens and latency for each stage of a run. **Who it is for:** A developer whose story is behaving oddly and who needs to see exactly what the model was asked and returned at each stage. **The problem:** Odd character behavior is hard to diagnose without seeing the prompts, outputs and cost behind each cognitive stage. **The outcome:** The user can find the stage and LLM call behind a specific story event, along with its prompt, response, tokens and latency. ### Steps 1. Run `./deploy/asamana.sh start` with the developer tools enabled 2. Run a world for a few steps so that LLM calls are recorded 3. Open the `#/dev` page and select the Trace view 4. Filter to the character or step that produced the odd behavior 5. Compare the prompt and response with the stage it belongs to, such as perception, motivation, decision, action or feedback ### Prompt to paste into your agent ```text In the repo https://github.com/dreamscroll-lab/asamana, run a world for 3 steps and open the Trace view. Report the stage name, token count and latency of the first recorded LLM call for one character. Verify that each recorded call is linked to a stage from the five-stage pipeline. ``` ### Caveat Trace data is shown in the developer tools, which the README says are for local development, so treat recorded prompts as potentially containing generated content. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/dreamscroll-lab-asamana/trace-llm-calls-to-debug-a-run/