Add knowledge graph

* migrate content from README.md to KG

README.md

@@ -62,122 +62,7 @@ docker run
 See [docs](https://vonwig.github.io/prompts.docs/#/page/running%20the%20prompt%20engine) for more details on how to run the conversation loop, 
 and especially how to use it to run local prompts that are not yet in GitHub.
 
-## Function volumes
-
-Every function container will have a shared volume mounted into the container at `/thread`.
-The volume is intended to be ephemeral and will be deleted at the end of the run.  However, the volume
-can be saved for inspection by passing the argument `--thread-id`.  
-
-## Output json-rpc notifications
-
-Add the flag `--jsonrpc` to the list of arguments to switch the stdout stream to be a series of `jsonrpc` notifications.  
-This is useful if you are running the tool and streaming responses on to a canvas.
-
-Try running the with the `--jsonrpc` to see a full example but the stdout stream will look something like this.
-
-```
-{"jsonrpc":"2.0","method":"message","params":{"content":" consistently"}}Content-Length: 65
-
-{"jsonrpc":"2.0","method":"message","params":{"content":" high"}}Content-Length: 61
-
-{"jsonrpc":"2.0","method":"message","params":{"content":"."}}Content-Length: 52
-
-{"jsonrpc":"2.0","method":"functions","params":null}Content-Length: 57
-
-{"jsonrpc":"2.0","method":"functions-done","params":null}Content-Length: 1703
-```
-
-### Notification Methods
-
-#### message
-
-This is a message from the assitant role, or from a tool role. 
-The params for the `message` method should be appended to the conversation.  The `params` can be either 
-`content` or `debug`.
-
-```json
-{"params": {"content": "append this output to the current message"}}
-{"params": {"debug": "this is a debug message and should only be shown in debug mode"}}
-```
-
-#### prompts
-
-Generated user and system prompts are sent to the client so that they can be displayed.  These
-are sent after extractors are expanded so that users can see the actual prompts sent to the AI model.
-
-```json
-{"params": {"messages": [{"role": "system", "content": "system prompt message"}]}}
-```
-
-#### functions
-
-Functions are json encoded strings.  When streaming, the content of the json params will change as 
-the functions streams.  This can be rendered in place to show the function definition completing
-as it streams.
-
-```json
-{"params": "{}"}
-```
-
-#### functions-done
-
-This notification is sent when a function definition has stopped streaming, and is now being executed.  
-The next notification after this will be a tool message.
-
-```json
-{"params": ""}
-```
-
-#### error
-
-The `error` notification is not a message from the model, prompts, or tools.  Instead, it represents a kind
-of _system_ error trying to run the conversation loop.  It should always be shown to the user as it 
-probably represents something like a networking error or a configuration problem.
-
-```json
-{"params": {"content": "error message"}}
-```
-
-### Request Methods
-
-#### prompt
-
-Send a user prompt into the converstation loop.  The `prompt` method takes the following `params`.
-
-```json
-{"params": {"content": "here is the user prompt"}}
-```
-
-## Prompt file layout
-
-Each prompt directory should contain a README.md describing the prompts and their purpose.  Each prompt file
-is a markdown document that supports moustache templates for subsituting context extracted from the project.
-
-```
-prompt_dir/
-├── 010_system_prompt.md
-├── 020_user_prompt.md
-└── README.md
-```
-
-* ordering of messages is determined by filename sorting
-* the role is encoded in the name of the file
-
-### Moustache Templates
-
-The prompt templates can contain expressions like {{dockerfiles}} to add information
-extracted from the current project.  Examples of facts that can be added to the
-prompts are:
-
-* `{{platform}}` - the platform of the current development environment.
-* `{{username}}` - the DockerHub username (and default namespace for image pushes)
-* `{{languages}}` - names of languages discovered in the project.
-* `{{project.dockerfiles}}` - the relative paths to local DockerFiles
-* `{{project.composefiles}}` - the relative paths to local Docker Compose files.
-
-The entire `project-facts` map is also available using dot-syntax
-forms like `{{project-facts.project-root-uri}}`.  All moustache template
-expressions documented [here](https://github.com/yogthos/Selmer) are supported.
+[PROMPTS KNOWLEDGE GRAPH](https://vonwig.github.io/prompts.docs/#/page/index)
 
 ## Building
 

graphs/prompts/journals/2024_09_03.md

@@ -0,0 +1,88 @@
+- Run some prompts checked in to GitHub against a project in the current working directory.
+  id:: 66d779c7-c1b7-40c6-a635-fa712da492de
+  ```sh
+  docker run 
+    --rm -it \
+    --pull=always \
+    -v /var/run/docker.sock:/var/run/docker.sock \
+    --mount type=volume,source=docker-prompts,target=/prompts \
+    --mount type=bind,source=$HOME/.openai-api-key,target=/root/.openai-api-key \
+    vonwig/prompts:latest \
+      run \
+      --host-dir $PWD \
+      --user jimclark106 \
+      --platform darwin \
+      --prompts "github:docker/labs-make-runbook?ref=main&path=prompts/lazy_docker"
+  ```
+	- Most of this is boiler plate except:
+		- the `--user` option in line 10 requires a valid DOCKER_HUB user name
+		- the `--prompts` option in line 12 requires a valid [github reference]([[GitHub Refs]]) to some markdown prompts
+		- if the project is located somewhere other than $PWD then the `--host-dir` will need to be updated.
+- Run a local prompt markdown file against a project in the current working directory.  In this example, the prompts are not pulled from GitHub.  Instead, our prompts are being developed in a directory called `$PROMPTS_DIR`.  In this example, the local prompt file is `$PROMPTS_DIR/myprompts.md`.
+  id:: 66d77f1b-1684-480d-ad7b-5e9f53292fe4
+  ```sh
+  docker run 
+    --rm -it \
+    --pull=always \
+    -v /var/run/docker.sock:/var/run/docker.sock \
+    --mount type=volume,source=docker-prompts,target=/prompts \
+    --mount type=bind,source=$HOME/.openai-api-key,target=/root/.openai-api-key \
+    --mount type=bind,source=$PROMPTS_DIR,target=/app/workdir \
+    --workdir /app/workdir \
+    vonwig/prompts:latest \
+      run \
+      --host-dir $PWD \
+      --user jimclark106 \
+      --platform darwin \
+      --prompts-file myprompts.md
+  ```
+	- Most of this is boiler plate except:
+		- the `--user` option in line 12 requires a valid DOCKER_HUB user name
+		- the `--prompts-file` option in line 14 is a relative path to a prompts file (relative to $PROMPTS_DIR)
+		- if the project being analyzed is located somewhere other than $PWD then the `--host-dir` will need to be updated.
+- [[Running the Prompt Engine]]
+- [[Authoring Prompts]]
+- Here is a prompt file with lots of non-default metadata (it uses  [extractors]([[Prompt Extractors]]), a [[tool]], and uses a local llm in [[ollama]]).  It uses one system prompt, and one user prompt.  Note that the user prompt contains a moustache template to pull data in from an extractor.
+  id:: 66d7f3ff-8769-40b3-b6b5-fc4fceea879e
+
+  ```md
+  ---
+  extractors:
+    - name: linguist
+      image: vonwig/go-linguist:latest
+      command:
+      - -json
+      output-handler: linguist
+  tools:
+    - name: findutils-by-name
+      description: find files in a project by name
+      parameters:
+          type: object
+          properties:
+            glob:
+              type: string
+              description: the glob pattern for files that should be found
+      container:
+          image: vonwig/findutils:latest
+          command:
+            - find
+            - .
+            - -name
+  model: llama3.1
+  url:  http://localhost/v1/chat/completions
+  stream: false
+  ---
+
+  # Prompt system
+
+  You are an expert on analyzing project content.
+
+  # Prompt user
+
+  {{#linguist}}
+  This project contains {{language}} code.
+  {{/linguist}}
+
+  Can you find any language specific project files and list them?
+
+  ```

graphs/prompts/journals/2024_09_04.md

@@ -0,0 +1,45 @@
+- An extractor is a function that runs before prompts are sent to an LLM.  It can _extract_ some data from a project directory in order inject context into a set of prompts.
+  id:: 66d87dd3-efa2-4eb3-ba92-5cc4c2f9700b
+- Create a docker container that expects a project bind mounted at `/project` and that writes `application/json` encoded data to `stdout`.  The data written to `stdout` is what will be made available to any subsequent prompt templates.  
+  id:: 66d8a36a-432f-4d1a-a48c-edbe0224b182
+  To test your extractor function, run
+  ```sh
+  docker run \ 
+    --rm -it \
+    --mount type=bind,source=$PWD,target=/project \
+    --workdir /project \
+    image-name:latest arg1 arg2 ....
+  ```
+	- this would make your current working directory available to the extractor at `/project`
+	- you can also arrange to pass arguments to the extractor function when you define the extractor metadata
+- Once you have defined an extractor image (eg `image-name:latest`), create an entry in the prompt file to reference it.
+  id:: 66d8a4f3-656d-42bf-b22a-60bba2d1887f
+  ```
+  ---
+  extractors:
+    - name: my-extractor
+      image: image-name:latest
+      command:
+      - arg1
+      - arg2
+  ---
+  
+  # Prompt user
+  
+  I can now inject context into the prompt using moustache template syntax.
+  {{#my-extractor}}
+  {{.}}
+  {{/my-extractor}}
+  ```
+  Read more about [moustache templates](https://mustache.github.io/mustache.5.html)
+- #log working on [[Prompt Extractors]]
+  #log working on [[Authoring Prompts]]
+- A very simple prompt file that contains no metadata, and just a single user prompt is
+  id:: 66d8a396-9268-4917-882f-da4c52b7b5dd
+  ```
+  # Prompt user
+  
+  Tell me about Docker!
+  ```
+- #log working on [[GitHub Refs]]
+-

graphs/prompts/journals/2024_09_05.md

@@ -0,0 +1,4 @@
+- [[conversation loop]]
+  id:: 66d9d1e0-a13e-4d62-8db7-9eebb37714a8
+-
+-