add doc on new cli commands

Signed-off-by: Sylvain Hellegouarch <sh@defuze.org>

Author: Lawouach

docs/fault/docs/explanations/fault-injection-basics.md

@@ -70,6 +70,22 @@ Each fault type has a distinct role in helping you simulate and analyze adverse
 - **Engineering Focus:**  
   Strengthen error-handling routines, validate user-friendly error messages, and implement effective retry or fallback mechanisms.
 
+### LLM Faults
+- **Purpose:**  
+  To adjust the parameters of LLM exchanges.
+- **Use Case:**  
+  Verify your application can handle unexpected responses from LLM.  
+- **Engineering Focus:**  
+  Strengthen error-handling routines, validate user-friendly error messages, and properly deal with LLM changes.
+
+### Database Faults
+- **Purpose:**  
+  To impact communications to and from databases.
+- **Use Case:**  
+  Explore how your application performs under database failures.  
+- **Engineering Focus:**  
+  Strengthen error-handling routines, validate user-friendly error messages, and properly deal with db errors.
+
 ## In Summary
 
 Fault injection is a powerful tool in your reliability engineering toolkit. It not only helps you detect vulnerabilities but also guides you in making informed improvements. By understanding the purpose behind each fault type and how to apply different distribution models, you can build robust systems that continue to perform even under duress.

docs/fault/docs/explanations/why-fault.md

@@ -23,9 +23,10 @@ systems lead to healthier operations.
 traffic through its proxy and test your application as normal:
 
 - Forward and tunnel proxy modes
-- HTTP and HTTPS
-- HTTP/1.1 and HTTP/2
-- TCP IPv4 transparent proxy
+- HTTP, HTTPS
+- HTTP/1.1, HTTP/2, SSE
+- TCP transparent proxy
+- LLM & Database high-level faults
 - Scenarii automation
 - eBPF stealth redirection on Linux
 

docs/fault/docs/how-to/install.md

@@ -6,9 +6,10 @@ on your environment.
 
 ## Features Matrix
 
-From a very high-level, <span class="f">fault</span> provides the following features:
+From a very high-level <span class="f">fault</span> provides the following features:
 
 * **Proxy**: a network proxy that model network traffic based on a configuration
+    - **LLM/DB**: proxy subcommands dedicated to explore LLM and database issues
 * **Scenario**: testing automation using the proxy
 * **Injection**: machinery to inject the network proxy into platform resources
 * **AI Agent**: review of results and code from a reliability and resilience perspective

docs/fault/docs/overrides/landing.html

@@ -408,23 +408,17 @@ <h3>Everything you need to know about bringing fault into your organization.</h3
         <div class="col-6">
           <details class="quote">
             <summary>What is fault?</summary>
-            <p>fault is a developer product aiming at supporting engineers keep their high-standards while onboarding AI
-              into their pipeline.</p>
+            <p>fault is a developer product aiming at supporting engineers keep their high-standards.</p>
           </details>
           <details class="quote">
             <summary>How is fault delivered?</summary>
-            <p>fault is a rust-cli. It runs natively on Linux, macOSX and Windows.</p>
+            <p>fault is a Rust-powered command line. It runs natively on Linux, macOSX and Windows.</p>
           </details>
           <details class="quote">
-            <summary>Is fault free?</summary>
-            <p>The fault CLI is free and open-source and will remain so. You will pay for any LLM model through your own
+            <summary>Is fault open-source?</summary>
+            <p>The fault CLI is free and open-source, using an Apache license, and will remain so. You will pay for any LLM model through your own
               subscription.</p>
           </details>
-          <details class="quote">
-            <summary>Do you offer enterprise commercial support?</summary>
-            <p>As part of the <a href="https://rebound.how/">Rebound</a> family, we do indeed support fault
-              commercially. Please feel free to <a href="https://rebound.how/support/#contact">reach out to us</a>.</p>
-          </details>
           <details class="quote">
             <summary>Do you upload my data anywhere?</summary>
             <p>The fault CLI doesn't send your data or code anywhere. If you use a Cloud-based LLM such as OpenAI,
@@ -488,10 +482,10 @@ <h2 class="footer-logo">
           </ul>
         </div>
         <div class="col-2">
-          <header>SUPPORT</header>
+          <header>COMMUNITY</header>
           <ul>
             <li><a href="https://github.com/rebound-how/rebound/issues">Issues</a></li>
-            <li><a href="https://github.com/orgs/rebound-how/discussions/categories/ideas">Feature Requests</a></li>
+            <li><a href="https://github.com/orgs/rebound-how/discussions/categories/ideas">Discussion</a></li>
             <li><a href="#faq">FAQ</a></li>
             <li><a href="https://rebound.how/support/#contact">Contact-Us</a></li>
           </ul>

docs/fault/docs/reference/cli-commands.md

@@ -12,7 +12,9 @@ defined in a file or launch a local demo server.
 ### `run`
 
 Run the proxy with fault injection enabled. This command applies the specified
-network faults to HTTP requests and tunnel streams.
+network faults to TCP streams and HTTP requests.
+
+It has two subcommands to specifically explore LLM and database use-cases.
 
 ### `inject`
 
@@ -354,12 +356,67 @@ Learn more about the [Blackhole fault](./builtin-faults.md#blackhole).
   _Default:_ `ingress`
 
 - **`--blackhole-sched <value>`**  
-  [Intervals scheduling](./schedule-intervals-syntax.md) when to apply the fault (require `--duration` whhen using relative schedule).  
+  [Intervals scheduling](./schedule-intervals-syntax.md) when to apply the fault (require `--duration` when using relative schedule).  
   **Example:** `--blackhole-sched "start:30s,duration:60s"`
   **Example:** `--blackhole-sched "start:5%,duration:40%"` (requires `--duration`)
 
 ---
 
+### `llm` Subcommand Options
+
+Specific faults to target your LLM.
+
+**<TARGET>**  Which LLM provider to target, one of `openai`, `gemini`,
+`open-router` and `ollama`
+
+- **`--endpoint`**
+  The base URL of the targeted LLM provider. Usually, you do not need to set
+  this value as the right base url will be set for each provider.
+
+- **`--case`**  
+  Which scenarios to run. Possible values `slow-stream`, `prompt-scramble`,
+  `token-drop`, `inject-bias`, `truncate-response`, `http-error`
+
+- **`--probability`**  
+    Fault injection probability between 0.0 (never) to 1.0 (always)
+  _Default:_ `1.0`
+
+Each case has its own parameters:
+
+When `--case` is `slow-stream`.
+
+- **`--slow-stream-mean-delay`**
+  Delay in miliseconds to slow the stream by.
+  _Default:_ `300`
+
+When `--case` is `token-drop`.
+
+No extra parameters.
+
+When `--case` is `prompt-scramble`.
+
+- **`--scramble-pattern`**
+  Optional regex pattern to scramble in prompt.
+
+- **`--scramble-with`**
+  Optional substitute text for scramble (must be set when `--scramble-pattern`
+  is set)
+
+- **`--instruction`**
+  Optional instruction/System prompt to set on the request.
+
+When `--case` is `inject-bias`.
+
+- **`--bias-pattern`**
+  Regex pattern for bias.
+
+- **`--bias-replacement`**
+    Substitute text for bias.
+
+When `--case` is `http-error`.
+
+No extra parameters.
+
 ### Usage Examples
 
 #### Running the Proxy with Multiple Faults
@@ -371,6 +428,13 @@ fault run \
   --with-bandwidth --bandwidth-rate 2000 --bandwidth-unit KBps
 ```
 
+#### Adding instructions to a LLM call
+
+```bash
+fault run llm openai --instruction "Respond in French"
+```
+
+
 ## `injection` Command Options
 
 Inject <span class="f">fault</span> into your platform resources.

docs/fault/docs/reference/environment-variables.md

@@ -66,6 +66,19 @@ be simpler to populate these options via environment variables.
 | `FAULT_DNS_PROBABILITY`                       | `0.5`                | Probability (0–100) to trigger a DNS fault.                                                    |
 | `FAULT_DNS_SCHED`                       | (none)                | Scheduling of the dns fault.                                                    |
 
+### `run llm` Command Variables
+
+| **Name**                         | **Default Value**   | **Explanation**                                                                                  |
+|----------------------------------|---------------------|--------------------------------------------------------------------------------------------------|
+| `FAULT_LLM_ENDPOINT`                  | (none)    | Base URL of the target LLM provider.                                                 |
+| `FAULT_LLM_PROBABILITY`                  | `1.0`    | Probability which will trigger the fault injection (0  means never and 1 means always).                                                 |
+| `FAULT_LLM_SLOW_STREAM_MEAN_DELAY`                  | `300`    | Latency to apply to the LLM response.                                                 |
+| `FAULT_LLM_SCRAMBLE_PATTERN`                  | (none)    | Regex pattern to look for into the request.                                                 |
+| `FAULT_LLM_SCRAMBLE_WITH`                  | (none)    | Replacement string when the pattern matches.                                                 |
+| `FAULT_LLM_SCRAMBLE_INSTRUCTION`                  | (none)    | Instruction to inject into the LLM requests as a system prompt.                                                 |
+| `FAULT_LLM_BIAS_PATTERN`                  | (none)    |  Regex pattern to look for into the response.                                                 |
+| `FAULT_LLM_BIAS_REPLACEMENT`                  | (none)    | Replacement string when the pattern matches.                                                 |
+
 ## `injection` Command Variables
 
 ### `aws` Subcommand Variables