add trigger docs + updates

apps/docusaurus/zeusfyi/docs/mockingbird/actions/_category_.json

@@ -0,0 +1,8 @@
+{
+  "label": "Actions",
+  "position": 4,
+  "link": {
+    "type": "generated-index",
+    "description": "How to build a manage mockingbird triggered actions."
+  }
+}

apps/docusaurus/zeusfyi/docs/mockingbird/actions/actions.md

@@ -0,0 +1,78 @@
+---
+sidebar_position: 2
+displayed_sidebar: mockingbird
+---
+
+# API Approvals
+
+## Overview
+
+The Actions tab is an essential feature for overseeing and managing API trigger calls within your workflow.
+It serves as a control panel where users can approve or reject pending actions, as well as view the historical
+results and responses of past actions. This ensures that all external communications via API calls are vetted by human
+oversight.
+
+## Approving and Rejecting API Triggers
+
+### Action Approval Process
+
+#### Review Request Summary:
+
+Examine the details of the API call request, including the Workflow Result ID, Approval ID, and the Request Summary.
+
+#### Approve or Reject:
+
+Make an informed decision to approve or reject the trigger action. Approving an action will execute the API call, while
+rejecting it will prevent the call from being made.
+
+### Update Status:
+
+Once an action is approved or rejected, its status is updated to reflect the new state.
+
+## Action Buttons
+
+### Approve:
+
+Click this button to authorize the API call.
+
+### Reject:
+
+Click this button to decline the API call.
+
+## Historical Actions
+
+### Viewing Past Actions
+
+### Access Historical Data:
+
+The lower section of the Actions tab displays a history of all API trigger actions that have been taken.
+
+Analyse Past Decisions: Review the Approval ID, Request Summary, Final State, Updated At, and the full list of requests
+and responses to understand the outcome of past decisions.
+
+### Historical Insights
+
+Gain insights into the decision-making process over time.
+Assess the consistency and accuracy of past approvals or rejections.
+
+### Example of the Actions Tab
+
+    Trigger Name: social-media
+    Trigger Group: social-media-approvals  
+    Trigger Env: social-media-engagement
+
+Action Details:
+
+- Workflow Result ID: 170665375241213000
+- Approval ID: 1706653756335363475
+- Approval State: Pending
+- Request Summary: Requesting approval for trigger action [...]
+
+Historical Actions:
+
+- Approval ID: 170664805335778953
+- Request Summary: Finished approval for trigger action
+- Final State: Finished
+- Updated At: 1/30/2024, 10:00:05 PM
+
+![Scre](https://github.com/zeus-fyi/zeus/assets/17446735/c1451545-4ea7-4c32-9415-01d079b43b39)

apps/docusaurus/zeusfyi/docs/mockingbird/api/_category_.json

@@ -0,0 +1,8 @@
+{
+  "label": "API",
+  "position": 5,
+  "link": {
+    "type": "generated-index",
+    "description": "How to use the mockingbird API."
+  }
+}

apps/docusaurus/zeusfyi/docs/mockingbird/api/api.md

@@ -0,0 +1,9 @@
+---
+sidebar_position: 2
+displayed_sidebar: mockingbird
+---
+
+# Docs
+
+### Coming Soon (Expected Mid-Feb 2024)
+

apps/docusaurus/zeusfyi/docs/mockingbird/builder/aggregations.md

@@ -9,4 +9,38 @@ Use this to tell the AI how to aggregate the results of your analysis chunks int
 aggregating on a single analysis, the aggregation cycle count sets how many base analysis cycles to aggregate on. If
 aggregating on multiple analysis, it will aggregate whenever the the underlying analysis is run.
 
+#### Diagram: Analysis->Aggregation Flow
+
+```mermaid
+flowchart TD
+    A[Analysis Stage] -->|Completion Choices| B[Aggregation Stage]
+    B --> C{Aggregation Decision}
+    C -->|Single Analysis| D[Aggregate Once]
+    C -->|Multiple Analyses| E[Wait for All Analyses]
+    E --> F[All Completed]
+    F --> G[Time Window Check]
+    G -->|Within Window| H[Aggregate Results]
+```
+
+## Text Aggregations
+
 ![gg](https://github.com/zeus-fyi/zeus/assets/17446735/7fa3a1f2-cf92-4442-8b7f-50841dca7f02)
+
+## JSON Aggregations
+
+![ScreenshotPM](https://github.com/zeus-fyi/zeus/assets/17446735/e7d816a2-439e-40e9-b732-4dacfe27e09d)
+
+### Example Run: Analysis->Aggregation Flows
+
+This shows an analysis stage feeding the aggregation stage. Notice how the Completion Choices output from the
+analysis stage are then used in the aggregation stage as the prompt body, which is how the aggregation stages work.
+By default it will use the last analysis stage in the workflow, relative to the aggregation stage, and
+for multiple analysis stages it will aggregate once every analysis stage has completed at least once
+and then aggregate all the results from the analysis stages that are within the time window since last aggregation.
+
+### Start::End SQL Search Window
+
+The start and end fields in the report are the generated SQL window query that the AI uses to aggregate the results
+when fetching from a time series indexed datastore.
+
+![Screens](https://github.com/zeus-fyi/zeus/assets/17446735/da52e86d-d24b-4ee0-9f49-37e2c3b6e7a8)

apps/docusaurus/zeusfyi/docs/mockingbird/builder/analysis.md

@@ -11,4 +11,16 @@ pieces and aggregate them into a final analysis result. Truncate will simply tru
 length it can support. If you set the max tokens field greater than 0, it becomes the maximum number of tokens to spend
 per task request.
 
-![zzz](https://github.com/zeus-fyi/zeus/assets/17446735/8340c438-5b19-4684-9b19-1cb0c973a5fe)
+![zzz](https://github.com/zeus-fyi/zeus/assets/17446735/8340c438-5b19-4684-9b19-1cb0c973a5fe)
+
+![Screens](https://github.com/zeus-fyi/zeus/assets/17446735/da52e86d-d24b-4ee0-9f49-37e2c3b6e7a8)
+
+### Example Run: Analysis->Aggregation Flows
+
+This shows an analysis stage feeding the aggregation stage. Notice how the Completion Choices output from the
+analysis stage are then used in the aggregation stage as the prompt body, which is how the aggregation stages work.
+By default it will use the last analysis stage in the workflow, relative to the aggregation stage, and
+for multiple analysis stages it will aggregate once every analysis stage has completed at least once
+and then aggregate all the results from the analysis stages that are within the time window since last aggregation.
+
+![Screens](https://github.com/zeus-fyi/zeus/assets/17446735/da52e86d-d24b-4ee0-9f49-37e2c3b6e7a8)

apps/docusaurus/zeusfyi/docs/mockingbird/builder/intro.md

@@ -9,3 +9,66 @@ The builder section allows you to build tasks (analysis, aggregation), retrieval
 then mix and match them to quickly create new workflows, which is ideal for rapid prototyping and experimentation.
 We tried to keep things as simple as possible to use via the UI to make it as accessible as possible. Let us know if
 anything is confusing or you have any questions.
+
+## Key Components of the Workflow Builder
+
+### Retrievals
+
+Purpose: This stage is responsible for the initial data gathering. Users define the retrieval by naming it and
+specifying the group and platform it associates with.
+Functionality: Set up multiple retrieval stages to collect data from different sources or services like databases, APIs,
+or social media platforms.
+
+### Analysis
+
+Function: Once data is retrieved, the analysis stage interprets and processes the information using selected AI models.
+Configuration: Assign a name, group, and the AI model to be used for analysis, along with setting the analysis cycle
+count, determining how frequently the analysis should occur.
+Aggregation
+
+### Aggregation
+
+Aggregation stages combine the results from multiple analysis cycles into a cohesive summary.
+Execution Logic: Define aggregation cycles relative to analysis cycles, ensuring aggregation happens after all necessary
+analyses are completed.
+Automated Evals
+Purpose: Automated evals are used to evaluate the results of analyses or aggregations against a set of criteria to
+validate or trigger further actions.
+Cycles: Configure eval cycles in tandem with analysis cycles to ensure evaluations are timely and relevant.
+
+### Fundamental Time Period
+
+Definition: This is the base time period against which all cycles are referenced, determining the workflow's execution
+intervals.
+Application: Set the fundamental time step and unit to establish how often the workflow runs, which can range from
+minutes to hours, depending on the needs of the task.
+
+## Workflow Builder View
+
+![Screen](https://github.com/zeus-fyi/zeus/assets/17446735/15ad98ff-ed0c-482c-9ced-9f4c08f222f7)
+
+## Creating a Workflow
+
+Define Workflow Name and Group: Start by giving your workflow a distinctive name and group for organizational purposes.
+
+### Add Stages
+
+Sequentially add retrieval, analysis, aggregation, and eval stages, specifying each stage's parameters and settings.
+
+### Connect Stages
+
+Link retrieval stages to analysis, ensuring data flows correctly through the workflow.
+
+### Set Time Periods
+
+Adjust the fundamental time period to control the overall pacing of the workflow execution.
+
+### Save Workflow
+
+Once all stages are correctly configured, save the workflow to activate it within the system.
+
+## Example Usage
+
+Consider a scenario where you are monitoring social media engagement. You would set up retrieval stages to pull data
+from social media APIs, analyze the sentiment and relevance using an AI model, aggregate the analysis results to track
+engagement trends over time, and evaluate these trends to decide on marketing strategies.

apps/docusaurus/zeusfyi/docs/mockingbird/builder/json_schemas.md

@@ -1,5 +1,5 @@
 ---
-sidebar_position: 7
+sidebar_position: 8
 displayed_sidebar: mockingbird
 ---
 

apps/docusaurus/zeusfyi/docs/mockingbird/builder/retrievals.md

@@ -3,18 +3,18 @@ sidebar_position: 5
 displayed_sidebar: mockingbird
 ---
 
-# Retrievals
+# Inputs/Outputs
 
 For generic retrievals we recommend you setup an API that can be called from the load balancer group for quick starting.
 
-## Web APIs via Load Balancer
+## APIs via Load Balancer
 
 It will get data from each route in the load balancer group you reference here and then aggregate the results into a
 single response.
 
-![Sc](https://github.com/zeus-fyi/zeus/assets/17446735/6c9f79e8-4ff7-4fc3-bc57-f4d425dbff46)
+![ScreensM](https://github.com/zeus-fyi/zeus/assets/17446735/f8011006-e92c-413a-9267-ea69165b91e9)
 
 The website extraction and keywords feature is still in development, but coming soon you'll be able to use it, but you
-can
-save instructions for now and it will be available in the future for you to update your retrievals.
+can save instructions for now and it will be available in the future for you to update your retrievals.
 
+See the RPC Load Balancer Documentation for adding routes, and creating routing groups that you can use here.

apps/docusaurus/zeusfyi/docs/mockingbird/builder/triggers.md

@@ -0,0 +1,66 @@
+---
+sidebar_position: 7
+displayed_sidebar: mockingbird
+---
+
+# Triggers and Actions
+
+## Intro
+
+The Triggers and Actions system is a critical component of our automation framework.
+It links evaluation functions to API triggers to streamline the execution of AI-driven workflows.
+Upon evaluation, this system generates a request payload for user approval or rejection.
+If approved, the corresponding API call is executed. Rejected requests are archived and not executed.
+This mechanism ensures that automated decisions are overseen and validated by human judgment when necessary.
+
+![Scr](https://github.com/zeus-fyi/zeus/assets/17446735/893cb682-b82d-4cbc-9cdf-742e41f7142b)
+
+## Trigger Action Procedures
+
+![ScreensM](https://github.com/zeus-fyi/zeus/assets/17446735/f8011006-e92c-413a-9267-ea69165b91e9)
+
+### Setting Up Triggers
+
+- **Define a New Trigger**: Assign a name and group to a new trigger, identify the trigger source, and specify the
+  trigger action.
+- **Link to Eval Function**: Connect the trigger to an eval function that will analyze AI outputs to decide on API call
+  initiation.
+
+### API Trigger Settings
+
+- **Retry Mechanisms**: Configure the maximum retry attempts for API calls and set up backoff coefficients for retry
+  intervals.
+- **Save Trigger**: Save the configured trigger so it can start listening for outputs from the linked eval function.
+
+## Evaluation Functions and API Calls
+
+![ScreensM](https://github.com/zeus-fyi/zeus/assets/17446735/f8011006-e92c-413a-9267-ea69165b91e9)
+
+### Connecting Eval Functions
+
+- **Select the Eval Function**: Pick the appropriate evaluation function that will provide criteria for API call
+  triggers.
+
+### Automated Payload Creation
+
+- **Payload Generation**: Once the eval function criteria are met, the system automatically generates an API call
+  request payload.
+
+## Actions Menu
+
+![Scree](https://github.com/zeus-fyi/zeus/assets/17446735/8f340b39-5c42-4813-9ca2-7e94334cafdf)
+
+### Review and Decision
+
+- **Approve/Reject Payloads**: Use the actions menu to review generated payloads and make decisions to approve or reject
+  them.
+
+### Historical Tracking
+
+- **Action Records**: The system maintains a log of all actions taken, which can be accessed for review or auditing
+  purposes.
+
+## Conclusion
+
+This triggers and actions framework empowers users to have final control over automated processes, ensuring that every
+action taken by the AI is deliberate and in line with organizational standards.

apps/docusaurus/zeusfyi/docs/mockingbird/runs/_category_.json

@@ -1,6 +1,6 @@
 {
   "label": "Runs",
-  "position": 2,
+  "position": 3,
   "link": {
     "type": "generated-index",
     "description": "How to manage mockingbird workflow runs."

apps/docusaurus/zeusfyi/docs/mockingbird/runs/reports.md

@@ -1,5 +1,5 @@
 ---
-sidebar_position: 2
+sidebar_position: 3
 displayed_sidebar: mockingbird
 ---
 
@@ -9,3 +9,18 @@ Shows the outputs of each stage within the workflow. You can click on the expand
 mockingbird run.
 
 ![ddd](https://github.com/zeus-fyi/zeus/assets/17446735/c77d77c4-0023-47f1-8196-66fdd93eeefc)
+
+### Example Run: Eval Results
+
+![ScreenshotPM](https://github.com/zeus-fyi/zeus/assets/17446735/648dbda2-ce22-4eea-b698-5c012844d190)
+
+### Example Run: Analysis->Aggregation Flows
+
+This shows an analysis stage feeding the aggregation stage. Notice how the Completion Choices output from the
+analysis stage are then used in the aggregation stage as the prompt body, which is how the aggregation stages work.
+By default it will use the last analysis stage in the workflow, relative to the aggregation stage, and
+for multiple analysis stages it will aggregate once every analysis stage has completed at least once
+and then aggregate all the results from the analysis stages that are within the time window since last aggregation.
+
+![Screens](https://github.com/zeus-fyi/zeus/assets/17446735/da52e86d-d24b-4ee0-9f49-37e2c3b6e7a8)
+

apps/docusaurus/zeusfyi/docs/mockingbird/setup/_category_.json

@@ -0,0 +1,8 @@
+{
+  "label": "Setup",
+  "position": 1,
+  "link": {
+    "type": "generated-index",
+    "description": "How to manage mockingbird setup options."
+  }
+}