# AI use cases for Aifw (securetron-gh/aifw) Repo: [securetron-gh/aifw](https://github.com/securetron-gh/aifw) | Use cases: 4 | Page: https://dedupedai.com/use-cases/securetron-gh-aifw/ AI-FW is a self-hosted gateway that inspects, governs, routes, and audits prompts and responses between AI agents and the models they call. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Put a firewall in front of your existing LLM calls: https://dedupedai.com/use-cases/securetron-gh-aifw/firewall-existing-llm-calls/ - Keep a metadata-only audit trail of agent calls: https://dedupedai.com/use-cases/securetron-gh-aifw/metadata-only-audit-trail/ - Route Claude Code and Cursor traffic through the gateway: https://dedupedai.com/use-cases/securetron-gh-aifw/route-claude-code-cursor-through-gateway/ - Give each AI agent its own identity and API key: https://dedupedai.com/use-cases/securetron-gh-aifw/give-each-agent-own-identity/ ## How to put a firewall in front of your existing LLM calls with Aifw Repo: [securetron-gh/aifw](https://github.com/securetron-gh/aifw) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Run AI-FW with Docker Compose and send your app's model calls through it for inspection and logging. **Who it is for:** A developer whose app already calls an OpenAI-compatible or Anthropic API and needs prompts and responses inspected before they reach the model. **The problem:** Model calls go straight to the provider, so there is no inspection, routing control, or record of what agents sent and received. **The outcome:** Your app's model traffic passes through the AI-FW gateway, and each transaction shows up in the dashboard. ### Steps 1. Run `docker compose up -d` in the folder that contains docker-compose.yml. 2. Open the gateway at http://localhost and sign in as superadmin with the default login from the README, then change it. 3. Register your backend URLs and models in the AI-FW Model Inventory (see Configure models and keys in the docs). 4. Point your app's OpenAI-compatible or Anthropic base URL at the gateway instead of the provider. 5. Run `curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{"model":"gpt-4o","messages":[{"role":"user","content":"Hello!"}]}'` and check the transaction in the dashboard. (inferred, check the README) ### Prompt to paste into your agent ```text In the securetron-gh/aifw repo, run `docker compose up -d` in , then send one chat completion request through the gateway to . Report the HTTP status and the transaction ID shown in the dashboard, and verify the response is a valid chat completion. ``` ### Caveat The compose setup listens on port 80 while the README curl example uses port 8080, so use port 80 when running with compose. 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/securetron-gh-aifw/firewall-existing-llm-calls/ ## How to keep a metadata-only audit trail of agent calls with Aifw Repo: [securetron-gh/aifw](https://github.com/securetron-gh/aifw) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README Use the dashboard and audit logs to review what AI agents sent and received through the gateway. **Who it is for:** A security or compliance lead who must show a regulator or internal reviewer how AI agents used models. **The problem:** Without a central record of agent model calls, reviewers cannot reconstruct what happened or prove policy was followed. **The outcome:** Each agent call is recorded as a metadata-only audit event that can be reviewed in the dashboard and exported. ### Steps 1. Run `docker compose up -d` in the folder that contains docker-compose.yml. 2. Send a few test requests through the gateway so transactions are recorded. 3. Open the observability dashboard and review the recorded transactions. 4. Open the audit logs section and review the metadata-only events. 5. Configure the encrypted audit export to your SIEM over syslog with TLS, as described in the audit logs documentation. (inferred, check the README) ### Prompt to paste into your agent ```text In the securetron-gh/aifw repo, start the gateway with `docker compose up -d` and send three test requests through it. Report how many transactions appear in the audit logs for , and verify each entry contains metadata and not full prompt text. ``` ### Caveat The README lists encrypted SIEM export under Enterprise, so confirm which export options your edition includes before relying on them. 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/securetron-gh-aifw/metadata-only-audit-trail/ ## How to route Claude Code and Cursor traffic through the gateway with Aifw Repo: [securetron-gh/aifw](https://github.com/securetron-gh/aifw) | Difficulty: Advanced | Time: 60 minutes | Basis: From the README Send coding agent model calls through AI-FW so prompts are inspected and governed. **Who it is for:** A developer or engineering lead whose team uses Claude Code or Cursor and wants their prompts inspected. **The problem:** Coding agents send prompts and code context straight to model providers with no inspection or policy checks. **The outcome:** Claude Code or Cursor requests pass through the gateway, and their transactions appear in the dashboard. ### Steps 1. Run `docker compose up -d` in the folder that contains docker-compose.yml. 2. Open the Claude Code, Cursor and MCP tutorial in docs/tutorials/02-claude-code-cursor-mcp.mdx. 3. Set the Claude Code or Cursor model endpoint to the gateway address, following the tutorial's steps. (inferred, check the README) 4. Send one coding prompt from the agent and confirm it appears in the dashboard. ### Prompt to paste into your agent ```text In the securetron-gh/aifw repo, follow docs/tutorials/02-claude-code-cursor-mcp.mdx to point at the running gateway. Report whether one test prompt appears in the dashboard, and verify the agent still gets a response. ``` ### Caveat The README points to the tutorial for the client setup and does not show the exact configuration values, so read the tutorial before editing agent settings. Risk, credentials: Coding agent configs will hold gateway and model API keys, so store them in your secret manager and not in shared repo files. 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/securetron-gh-aifw/route-claude-code-cursor-through-gateway/ ## How to give each AI agent its own identity and API key with Aifw Repo: [securetron-gh/aifw](https://github.com/securetron-gh/aifw) | Difficulty: Advanced | Time: 60 minutes | Basis: From the README Register agents as identities with admin-issued API keys so their calls can be told apart. **Who it is for:** A platform admin running several AI agents who needs each one authorized and traceable on its own. **The problem:** Agents sharing one key cannot be told apart in logs and cannot be limited or revoked individually. **The outcome:** Each agent has its own registered identity and admin-issued key, and its calls are traceable in the dashboard. ### Steps 1. Run `docker compose up -d` in the folder that contains docker-compose.yml. 2. Sign in as superadmin at http://localhost with the default login, then change the password. 3. Open the agents section in the admin pages and register one agent identity per agent. (inferred, check the README) 4. Issue an admin-issued API key for each registered agent from the API keys page. (inferred, check the README) 5. Follow docs/tutorials/04-agent-self-enrollment.mdx if agents should enroll themselves instead. ### Prompt to paste into your agent ```text In the securetron-gh/aifw repo, start the gateway with `docker compose up -d` and register as its own identity with its own API key. Report the agent identity and key name, and verify that a test call made with that key shows up under that agent in the dashboard. ``` ### Caveat The README names the agent registry and admin-issued keys but does not show their exact admin steps, so the agent registration steps above are inferred from the docs layout. Risk, credentials: The README lists a default superadmin login, so change it before the gateway is reachable from any network, and store each agent key in a secret manager. 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/securetron-gh-aifw/give-each-agent-own-identity/