ultravisor 1.0.21 → 1.0.23

package/docs/features/tasks-llm.md

@@ -10,39 +10,39 @@ Sends messages to a large language model and returns the completion. Supports mu
 
 ### Settings
 
-- **SystemPrompt** — System prompt text that sets the LLM's behavior and context.
-- **UserPrompt** — User prompt text for the current turn.
-- **Messages** — JSON array of message objects `[{role, content}]` for direct control over the conversation. Overrides SystemPrompt/UserPrompt when set.
-- **Model** — Override the default model name.
-- **Temperature** — Sampling temperature (0–2). Lower values are more deterministic.
-- **MaxTokens** — Maximum tokens to generate (default `4096`).
-- **TopP** — Nucleus sampling parameter.
-- **StopSequences** — JSON array of sequences that stop generation.
-- **ResponseFormat** — Set to `json_object` to force JSON output.
-- **InputAddress** — State address to read additional context data from (appended to UserPrompt).
-- **Destination** — State address to write the completion text to.
+- **SystemPrompt** -- System prompt text that sets the LLM's behavior and context.
+- **UserPrompt** -- User prompt text for the current turn.
+- **Messages** -- JSON array of message objects `[{role, content}]` for direct control over the conversation. Overrides SystemPrompt/UserPrompt when set.
+- **Model** -- Override the default model name.
+- **Temperature** -- Sampling temperature (0-2). Lower values are more deterministic.
+- **MaxTokens** -- Maximum tokens to generate (default `4096`).
+- **TopP** -- Nucleus sampling parameter.
+- **StopSequences** -- JSON array of sequences that stop generation.
+- **ResponseFormat** -- Set to `json_object` to force JSON output.
+- **InputAddress** -- State address to read additional context data from (appended to UserPrompt).
+- **Destination** -- State address to write the completion text to.
 
 ### Conversation History
 
-- **ConversationAddress** — State address to store message history for multi-turn conversations.
-- **AppendToConversation** — Append this exchange to history (default `true` when ConversationAddress is set).
-- **ConversationMaxMessages** — Sliding window: maximum non-system messages to keep.
-- **ConversationMaxTokens** — Token budget for history (approximate, trims oldest messages).
-- **PersistConversation** — Copy conversation history to GlobalState for cross-operation persistence.
-- **ConversationPersistAddress** — GlobalState address for persistence.
+- **ConversationAddress** -- State address to store message history for multi-turn conversations.
+- **AppendToConversation** -- Append this exchange to history (default `true` when ConversationAddress is set).
+- **ConversationMaxMessages** -- Sliding window: maximum non-system messages to keep.
+- **ConversationMaxTokens** -- Token budget for history (approximate, trims oldest messages).
+- **PersistConversation** -- Copy conversation history to GlobalState for cross-operation persistence.
+- **ConversationPersistAddress** -- GlobalState address for persistence.
 
 ### Dispatch
 
-- **AffinityKey** — Route to a specific Beacon worker for consistent model access.
-- **TimeoutMs** — Timeout in milliseconds (default `120000`).
+- **AffinityKey** -- Route to a specific Beacon worker for consistent model access.
+- **TimeoutMs** -- Timeout in milliseconds (default `120000`).
 
 ### Outputs
 
-- **Content** — The LLM's response text.
-- **Model** — Model that generated the response.
-- **PromptTokens** / **CompletionTokens** / **TotalTokens** — Token usage statistics.
-- **FinishReason** — Why the completion ended (`stop`, `length`, etc.).
-- **BeaconID** — ID of the Beacon worker that executed the request.
+- **Content** -- The LLM's response text.
+- **Model** -- Model that generated the response.
+- **PromptTokens** / **CompletionTokens** / **TotalTokens** -- Token usage statistics.
+- **FinishReason** -- Why the completion ended (`stop`, `length`, etc.).
+- **BeaconID** -- ID of the Beacon worker that executed the request.
 
 ### Tips
 
@@ -56,20 +56,20 @@ Generates vector embeddings for text input using an LLM provider. Dispatches the
 
 ### Settings
 
-- **Text** — The text to embed. Supports Pict template expressions.
-- **Model** — Override the default embedding model.
-- **Dimensions** — Requested embedding dimensions (model-dependent).
-- **InputAddress** — State address to read text from (alternative to Text setting).
-- **Destination** — State address to write the embedding vector to.
-- **AffinityKey** — Route to a specific Beacon worker.
-- **TimeoutMs** — Timeout in milliseconds (default `60000`).
+- **Text** -- The text to embed. Supports Pict template expressions.
+- **Model** -- Override the default embedding model.
+- **Dimensions** -- Requested embedding dimensions (model-dependent).
+- **InputAddress** -- State address to read text from (alternative to Text setting).
+- **Destination** -- State address to write the embedding vector to.
+- **AffinityKey** -- Route to a specific Beacon worker.
+- **TimeoutMs** -- Timeout in milliseconds (default `60000`).
 
 ### Outputs
 
-- **Embedding** — JSON array of floating-point numbers representing the embedding vector.
-- **Dimensions** — Number of dimensions in the embedding.
-- **Model** — Model used for embedding.
-- **BeaconID** — ID of the Beacon that executed the work.
+- **Embedding** -- JSON array of floating-point numbers representing the embedding vector.
+- **Dimensions** -- Number of dimensions in the embedding.
+- **Model** -- Model used for embedding.
+- **BeaconID** -- ID of the Beacon that executed the work.
 
 ### Tips
 
@@ -83,35 +83,35 @@ Sends messages to a large language model with tool (function) definitions, enabl
 
 ### Settings
 
-- **SystemPrompt** — System prompt text.
-- **UserPrompt** — User prompt text.
-- **Messages** — JSON array of messages for direct conversation control.
-- **Tools** — JSON array of tool definitions describing available functions the LLM can call.
-- **Model** — Override model name.
-- **ToolChoice** — `auto` (LLM decides), `none` (no tools), or a specific tool name (default `auto`).
-- **Temperature** — Sampling temperature.
-- **MaxTokens** — Maximum tokens to generate.
-- **ConversationAddress** — State address for multi-turn history.
-- **AppendToConversation** — Append this exchange to history.
-- **InputAddress** — State address to read context data from.
-- **Destination** — State address to write completion content.
-- **AffinityKey** — Route to a specific Beacon worker.
-- **TimeoutMs** — Timeout in milliseconds (default `120000`).
+- **SystemPrompt** -- System prompt text.
+- **UserPrompt** -- User prompt text.
+- **Messages** -- JSON array of messages for direct conversation control.
+- **Tools** -- JSON array of tool definitions describing available functions the LLM can call.
+- **Model** -- Override model name.
+- **ToolChoice** -- `auto` (LLM decides), `none` (no tools), or a specific tool name (default `auto`).
+- **Temperature** -- Sampling temperature.
+- **MaxTokens** -- Maximum tokens to generate.
+- **ConversationAddress** -- State address for multi-turn history.
+- **AppendToConversation** -- Append this exchange to history.
+- **InputAddress** -- State address to read context data from.
+- **Destination** -- State address to write completion content.
+- **AffinityKey** -- Route to a specific Beacon worker.
+- **TimeoutMs** -- Timeout in milliseconds (default `120000`).
 
 ### Outputs
 
-- **Content** — The LLM's text response (may be empty if tool calls were made).
-- **ToolCalls** — JSON array of tool call objects with function names and arguments.
-- **Model** — Model that generated the response.
-- **FinishReason** — `stop` for normal completion, `tool_calls` when tools were invoked.
-- **PromptTokens** / **CompletionTokens** — Token usage.
-- **BeaconID** — Beacon worker ID.
+- **Content** -- The LLM's text response (may be empty if tool calls were made).
+- **ToolCalls** -- JSON array of tool call objects with function names and arguments.
+- **Model** -- Model that generated the response.
+- **FinishReason** -- `stop` for normal completion, `tool_calls` when tools were invoked.
+- **PromptTokens** / **CompletionTokens** -- Token usage.
+- **BeaconID** -- Beacon worker ID.
 
 ### Events
 
-- **Complete** — Fires when the LLM responds.
-- **ToolCall** — Fires when the LLM requests a tool call.
-- **Error** — Fires on failure.
+- **Complete** -- Fires when the LLM responds.
+- **ToolCall** -- Fires when the LLM requests a tool call.
+- **Error** -- Fires on failure.
 
 ### Tips
 

package/docs/features/tasks-meadow-api.md

@@ -10,20 +10,20 @@ Creates a new record via a Meadow REST API endpoint.
 
 ### Settings
 
-- **Entity** — Entity (table) name.
-- **Endpoint** — Base URL of the Meadow API server.
-- **DataAddress** — State address of the object containing the record data to create.
-- **Headers** — JSON string of request headers for authentication.
-- **Destination** — State address to store the created record (includes server-generated fields like ID).
+- **Entity** -- Entity (table) name.
+- **Endpoint** -- Base URL of the Meadow API server.
+- **DataAddress** -- State address of the object containing the record data to create.
+- **Headers** -- JSON string of request headers for authentication.
+- **Destination** -- State address to store the created record (includes server-generated fields like ID).
 
 ### Outputs
 
-- **Created** — The newly created record object with its assigned ID.
+- **Created** -- The newly created record object with its assigned ID.
 
 ### Events
 
-- **Complete** — Fires after the record is created.
-- **Error** — Fires on failure.
+- **Complete** -- Fires after the record is created.
+- **Error** -- Fires on failure.
 
 ### Tips
 
@@ -37,20 +37,20 @@ Reads a single record by its ID from a Meadow REST API endpoint.
 
 ### Settings
 
-- **Entity** — Entity (table) name to query.
-- **Endpoint** — Base URL of the Meadow API server.
-- **RecordID** — The ID of the record to retrieve.
-- **Destination** — State address to store the retrieved record.
-- **Headers** — JSON string of additional request headers for authentication.
+- **Entity** -- Entity (table) name to query.
+- **Endpoint** -- Base URL of the Meadow API server.
+- **RecordID** -- The ID of the record to retrieve.
+- **Destination** -- State address to store the retrieved record.
+- **Headers** -- JSON string of additional request headers for authentication.
 
 ### Outputs
 
-- **Record** — The retrieved record object.
+- **Record** -- The retrieved record object.
 
 ### Events
 
-- **Complete** — Fires after the record is retrieved.
-- **Error** — Fires if the record is not found or the request fails.
+- **Complete** -- Fires after the record is retrieved.
+- **Error** -- Fires if the record is not found or the request fails.
 
 ### Tips
 
@@ -64,23 +64,23 @@ Reads multiple records from a Meadow REST API endpoint, with optional filtering
 
 ### Settings
 
-- **Entity** — Entity (table) name to query.
-- **Endpoint** — Base URL of the Meadow API server.
-- **Filter** — Meadow filter expression to narrow the result set (e.g. `FBV~IDUser~EQ~42~0~`).
-- **Destination** — State address to store the records array.
-- **Headers** — JSON string of request headers for authentication.
-- **PageSize** — Number of records per page (default `100`).
-- **PageNumber** — Zero-based page number (default `0`).
+- **Entity** -- Entity (table) name to query.
+- **Endpoint** -- Base URL of the Meadow API server.
+- **Filter** -- Meadow filter expression to narrow the result set (e.g. `FBV~IDUser~EQ~42~0~`).
+- **Destination** -- State address to store the records array.
+- **Headers** -- JSON string of request headers for authentication.
+- **PageSize** -- Number of records per page (default `100`).
+- **PageNumber** -- Zero-based page number (default `0`).
 
 ### Outputs
 
-- **Records** — Array of retrieved record objects.
-- **RecordCount** — Number of records returned.
+- **Records** -- Array of retrieved record objects.
+- **RecordCount** -- Number of records returned.
 
 ### Events
 
-- **Complete** — Fires after records are retrieved.
-- **Error** — Fires on request failure.
+- **Complete** -- Fires after records are retrieved.
+- **Error** -- Fires on request failure.
 
 ### Tips
 
@@ -94,20 +94,20 @@ Updates an existing record via a Meadow REST API endpoint.
 
 ### Settings
 
-- **Entity** — Entity (table) name.
-- **Endpoint** — Base URL of the Meadow API server.
-- **DataAddress** — State address of the record data to update. Must include the record's ID field.
-- **Headers** — JSON string of request headers for authentication.
-- **Destination** — State address to store the updated record.
+- **Entity** -- Entity (table) name.
+- **Endpoint** -- Base URL of the Meadow API server.
+- **DataAddress** -- State address of the record data to update. Must include the record's ID field.
+- **Headers** -- JSON string of request headers for authentication.
+- **Destination** -- State address to store the updated record.
 
 ### Outputs
 
-- **Updated** — The updated record object.
+- **Updated** -- The updated record object.
 
 ### Events
 
-- **Complete** — Fires after the update.
-- **Error** — Fires on failure.
+- **Complete** -- Fires after the update.
+- **Error** -- Fires on failure.
 
 ### Tips
 
@@ -121,19 +121,19 @@ Deletes a record by its ID via a Meadow REST API endpoint.
 
 ### Settings
 
-- **Entity** — Entity (table) name.
-- **Endpoint** — Base URL of the Meadow API server.
-- **RecordID** — The ID of the record to delete.
-- **Headers** — JSON string of request headers for authentication.
+- **Entity** -- Entity (table) name.
+- **Endpoint** -- Base URL of the Meadow API server.
+- **RecordID** -- The ID of the record to delete.
+- **Headers** -- JSON string of request headers for authentication.
 
 ### Events
 
-- **Done** — Fires after the record is deleted.
-- **Error** — Fires on failure.
+- **Done** -- Fires after the record is deleted.
+- **Error** -- Fires on failure.
 
 ### Tips
 
-Meadow Delete is permanent — consider adding a **Value Input** confirmation step before deleting records in user-facing workflows. Build the RecordID dynamically with template expressions when deleting records identified by earlier processing steps.
+Meadow Delete is permanent -- consider adding a **Value Input** confirmation step before deleting records in user-facing workflows. Build the RecordID dynamically with template expressions when deleting records identified by earlier processing steps.
 
 ---
 
@@ -143,20 +143,20 @@ Counts records for an entity via a Meadow REST API endpoint, with optional filte
 
 ### Settings
 
-- **Entity** — Entity (table) name.
-- **Endpoint** — Base URL of the Meadow API server.
-- **Destination** — State address to store the count value.
-- **Headers** — JSON string of request headers for authentication.
-- **Filter** — Meadow filter expression to count only matching records.
+- **Entity** -- Entity (table) name.
+- **Endpoint** -- Base URL of the Meadow API server.
+- **Destination** -- State address to store the count value.
+- **Headers** -- JSON string of request headers for authentication.
+- **Filter** -- Meadow filter expression to count only matching records.
 
 ### Outputs
 
-- **Count** — Number of records matching the criteria.
+- **Count** -- Number of records matching the criteria.
 
 ### Events
 
-- **Complete** — Fires after the count is retrieved.
-- **Error** — Fires on failure.
+- **Complete** -- Fires after the count is retrieved.
+- **Error** -- Fires on failure.
 
 ### Tips
 

package/docs/features/tasks-user-interaction.md

@@ -10,20 +10,20 @@ Pauses flow execution and prompts the user for input. The flow resumes when the
 
 ### Settings
 
-- **PromptMessage** — The message displayed to the user (default: "Please provide a value:").
-- **OutputAddress** — State address where the user's input will be stored.
-- **InputType** — Type of input control: `text`, `number`, `boolean`, or `select` (default `text`).
-- **DefaultValue** — Pre-filled default value.
-- **Options** — JSON array of allowed values when InputType is `select`.
+- **PromptMessage** -- The message displayed to the user (default: "Please provide a value:").
+- **OutputAddress** -- State address where the user's input will be stored.
+- **InputType** -- Type of input control: `text`, `number`, `boolean`, or `select` (default `text`).
+- **DefaultValue** -- Pre-filled default value.
+- **Options** -- JSON array of allowed values when InputType is `select`.
 
 ### Outputs
 
-- **InputValue** — The value provided by the user.
+- **InputValue** -- The value provided by the user.
 
 ### Events
 
-- **ValueInputComplete** — Fires when the user submits a value.
-- **Cancelled** — Fires if the user cancels without providing input.
+- **ValueInputComplete** -- Fires when the user submits a value.
+- **Cancelled** -- Fires if the user cancels without providing input.
 
 ### Tips
 
@@ -37,13 +37,13 @@ Logs an error, warning, info, or debug message to the execution log. Use this ca
 
 ### Settings
 
-- **MessageTemplate** — The message text to log. Supports Pict template expressions for including state values (default: "An error occurred.").
-- **Level** — Log level: `error`, `warning`, `info`, or `debug` (default `error`).
+- **MessageTemplate** -- The message text to log. Supports Pict template expressions for including state values (default: "An error occurred.").
+- **Level** -- Log level: `error`, `warning`, `info`, or `debug` (default `error`).
 
 ### Events
 
-- **Complete** — Fires after the message is logged. Execution continues normally.
+- **Complete** -- Fires after the message is logged. Execution continues normally.
 
 ### Tips
 
-Error Message does not stop the flow — it only records a log entry. Place it on error branches to capture diagnostic context when failures occur. Use the `info` level for progress markers and `debug` for development-time tracing.
+Error Message does not stop the flow -- it only records a log entry. Place it on error branches to capture diagnostic context when failures occur. Use the `info` level for progress markers and `debug` for development-time tracing.

package/docs/features/tasks.md

@@ -8,15 +8,15 @@ supported task type.
 
 For detailed documentation on each built-in task type, see the per-category reference pages:
 
-- [Data Transform](tasks-data-transform.md) — Set values, string manipulation, templates, expressions, CSV parsing, and dataset operations
-- [File System](tasks-file-system.md) — Read, write, copy, and list files on the local file system
-- [Flow Control](tasks-flow-control.md) — Conditionals, loops, sub-operations, and shell commands
-- [HTTP Client](tasks-http-client.md) — GET/POST requests, JSON APIs, and fully configurable REST calls
-- [LLM](tasks-llm.md) — Chat completions, embeddings, and tool use with large language models
-- [Meadow API](tasks-meadow-api.md) — CRUD operations against Meadow REST API endpoints
-- [Content System](tasks-content-system.md) — Remote file operations on Content System beacon workers
-- [Extension](tasks-extension.md) — Low-level Beacon dispatch for remote worker execution
-- [User Interaction](tasks-user-interaction.md) — User input prompts and diagnostic logging
+- [Data Transform](tasks-data-transform.md) -- Set values, string manipulation, templates, expressions, CSV parsing, and dataset operations
+- [File System](tasks-file-system.md) -- Read, write, copy, and list files on the local file system
+- [Flow Control](tasks-flow-control.md) -- Conditionals, loops, sub-operations, and shell commands
+- [HTTP Client](tasks-http-client.md) -- GET/POST requests, JSON APIs, and fully configurable REST calls
+- [LLM](tasks-llm.md) -- Chat completions, embeddings, and tool use with large language models
+- [Meadow API](tasks-meadow-api.md) -- CRUD operations against Meadow REST API endpoints
+- [Content System](tasks-content-system.md) -- Remote file operations on Content System beacon workers
+- [Extension](tasks-extension.md) -- Low-level Beacon dispatch for remote worker execution
+- [User Interaction](tasks-user-interaction.md) -- User input prompts and diagnostic logging
 
 ## Task Model
 
@@ -82,8 +82,8 @@ is not set.
 ```
 
 Default execution limits (configurable in `.ultravisor.json`):
-- Timeout: 5 minutes (300,000 ms) — `UltravisorCommandTimeoutMilliseconds`
-- Max output buffer: 10 MB — `UltravisorCommandMaxBufferBytes`
+- Timeout: 5 minutes (300,000 ms) -- `UltravisorCommandTimeoutMilliseconds`
+- Max output buffer: 10 MB -- `UltravisorCommandMaxBufferBytes`
 
 ### Request
 
@@ -246,7 +246,7 @@ Reads a text file from the staging folder and returns its content in
 
 Writes an XML string to a file in the staging folder.
 Creates intermediate directories automatically. No XML validation is
-performed — the caller is responsible for providing well-formed XML.
+performed -- the caller is responsible for providing well-formed XML.
 
 | Field | Required | Description |
 |-------|----------|-------------|
@@ -270,7 +270,7 @@ the XML content is resolved from `pContext.GlobalState`.
 ### ReadXML
 
 Reads an XML file from the staging folder and returns its content as a
-raw string in `Output`. No XML parsing is performed — the caller is
+raw string in `Output`. No XML parsing is performed -- the caller is
 responsible for interpreting the XML structure.
 
 | Field | Required | Description |
@@ -376,7 +376,7 @@ Use `Address` to copy a file whose path was determined by a previous task:
 ```
 
 The source must be an existing regular file (not a directory). Path
-traversal is blocked in the destination — file paths containing `..`
+traversal is blocked in the destination -- file paths containing `..`
 are rejected.
 
 ### GetJSON
@@ -463,7 +463,7 @@ The raw response text is available in `pManifestEntry.Output`. The
 
 Performs a native HTTP/HTTPS GET request and returns the response body as
 raw XML text. Uses Node.js built-in `http`/`https` modules (no curl
-dependency). No XML parsing is performed — the caller is responsible
+dependency). No XML parsing is performed -- the caller is responsible
 for interpreting the XML structure.
 
 | Field | Required | Description |
@@ -572,7 +572,7 @@ The `JSON` field is only present when the response body is valid JSON.
 When a response includes `Set-Cookie` headers, the name/value pairs are
 automatically parsed and stored at `pContext.GlobalState.Cookies`. Every
 subsequent `RestRequest` task running in the same operation will
-automatically include those cookies in its request — no extra
+automatically include those cookies in its request -- no extra
 configuration needed.
 
 Task-level `Cookies` merge on top of the shared jar, so explicit values
@@ -688,7 +688,7 @@ name to store the value under):
 If the response body is `{"Session": {"ID": "xyz789"}}`, then
 `GlobalState.Cookies.SessionID` is set to `"xyz789"`.
 
-Nested paths are supported — the address walks the JSON structure
+Nested paths are supported -- the address walks the JSON structure
 using dot-separated keys.
 
 #### CaptureHeader
@@ -1267,9 +1267,9 @@ Tasks within a set execute sequentially in array order.
 ```
 1. onBefore[0], onBefore[1], ...     (always runs)
 2. Core task execution
-3. If success  → onCompletion[0], onCompletion[1], ...
-   If failure  → onFailure[0], onFailure[1], ...
-   If error    → onError[0], onError[1], ...
+3. If success  -> onCompletion[0], onCompletion[1], ...
+   If failure  -> onFailure[0], onFailure[1], ...
+   If error    -> onError[0], onError[1], ...
 4. onSubsequent[0], onSubsequent[1], ... (always runs)
 ```
 

package/docs/features/universal-addressing.md

@@ -8,10 +8,10 @@ Ultravisor's universal addressing system provides a single, human-readable schem
 >BeaconName/Context/Path/to/resource
 ```
 
-- **`>`** — Prefix indicating a universal address
-- **BeaconName** — Human-readable beacon name (e.g., `retold-remote`, `orator-conversion`)
-- **Context** — A named namespace on the beacon (e.g., `File`, `Cache`, `Staging`)
-- **Path** — Resource path within the context
+- **`>`** -- Prefix indicating a universal address
+- **BeaconName** -- Human-readable beacon name (e.g., `retold-remote`, `orator-conversion`)
+- **Context** -- A named namespace on the beacon (e.g., `File`, `Cache`, `Staging`)
+- **Path** -- Resource path within the context
 
 ### Examples
 
@@ -36,7 +36,7 @@ sequenceDiagram
     RA->>BC: resolveUniversalAddress(address)
     BC->>BC: Parse: beacon="retold-remote", context="File", path="photo.jpg"
     BC->>BC: findBeaconByName("retold-remote")
-    BC->>BC: Look up Context "File" → BaseURL
+    BC->>BC: Look up Context "File" -> BaseURL
     BC-->>RA: { URL, BeaconID, BeaconName, Context, Path, Filename }
 
     alt RequestingBeaconID provided
@@ -50,11 +50,11 @@ sequenceDiagram
 
 ### Resolution Steps
 
-1. **Parse** — Strip the `>` prefix, split into `BeaconName/Context/Path`
-2. **Beacon Lookup** — Find the beacon by name (or ID) in the coordinator's registry
-3. **Context Lookup** — Retrieve the context definition from the beacon record (contains `BaseURL`, `BasePath`, `Writable`, `Description`)
-4. **URL Construction** — Combine `BaseURL` with URL-encoded path segments
-5. **Strategy Resolution** (optional) — If `RequestingBeaconID` is provided, consult the reachability matrix to determine the best transfer strategy
+1. **Parse** -- Strip the `>` prefix, split into `BeaconName/Context/Path`
+2. **Beacon Lookup** -- Find the beacon by name (or ID) in the coordinator's registry
+3. **Context Lookup** -- Retrieve the context definition from the beacon record (contains `BaseURL`, `BasePath`, `Writable`, `Description`)
+4. **URL Construction** -- Combine `BaseURL` with URL-encoded path segments
+5. **Strategy Resolution** (optional) -- If `RequestingBeaconID` is provided, consult the reachability matrix to determine the best transfer strategy
 
 ## Contexts
 
@@ -79,7 +79,7 @@ beacon.registerContext('File', {
 
 ## Operations as Universal Data Locators
 
-Operation pipelines use universal addresses as their primary input mechanism. When a caller triggers an operation, it passes addresses as parameters — the operation graph handles resolution, transfer, processing, and result delivery automatically.
+Operation pipelines use universal addresses as their primary input mechanism. When a caller triggers an operation, it passes addresses as parameters -- the operation graph handles resolution, transfer, processing, and result delivery automatically.
 
 ```mermaid
 graph LR
@@ -90,7 +90,7 @@ graph LR
     E -->|Binary stream| F[Caller]
 ```
 
-This decouples the caller from knowing where files are or how to reach them. The address `>retold-remote/File/photo.jpg` works whether the beacon is on localhost, across a LAN, or behind a proxy — the resolve-address card and reachability matrix handle the routing.
+This decouples the caller from knowing where files are or how to reach them. The address `>retold-remote/File/photo.jpg` works whether the beacon is on localhost, across a LAN, or behind a proxy -- the resolve-address card and reachability matrix handle the routing.
 
 ## resolve-address Card Reference
 

package/docs/index.html

@@ -4,9 +4,9 @@
 		<meta charset="utf-8">
 		<meta http-equiv="X-UA-Compatible" content="IE=edge">
 		<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
-		<meta name="description" content="Documentation powered by pict-docuserve">
+		<meta name="description" content="Ultravisor v1.0.21 Documentation — Cyclic process execution with ai integration.">
 
-		<title>Documentation</title>
+		<title>Ultravisor v1.0.21 Documentation</title>
 
 		<!-- Application Stylesheet -->
 		<link href="css/docuserve.css" rel="stylesheet">

package/docs/retold-catalog.json

@@ -1,5 +1,5 @@
 {
-	"Generated": "2026-03-21T22:11:31.751Z",
+	"Generated": "2026-04-10T17:13:49.181Z",
 	"GitHubOrg": "stevenvelozo",
 	"DefaultBranch": "master",
 	"Groups": [
@@ -263,6 +263,22 @@
 									"Title": "Building Beacon Providers",
 									"Path": "features/beacon-providers.md"
 								},
+								{
+									"Title": "Beacon Authentication",
+									"Path": "features/beacon-authentication.md"
+								},
+								{
+									"Title": "Universal Addressing",
+									"Path": "features/universal-addressing.md"
+								},
+								{
+									"Title": "Reachability Matrix",
+									"Path": "features/reachability-matrix.md"
+								},
+								{
+									"Title": "Platform Cards",
+									"Path": "features/platform-cards.md"
+								},
 								{
 									"Title": "LLM Integration",
 									"Path": "features/llm.md"
@@ -342,6 +358,15 @@
 									"Path": "features/configuration.md"
 								}
 							]
+						},
+						{
+							"Title": "Case Studies",
+							"Children": [
+								{
+									"Title": "retold-remote + orator-conversion",
+									"Path": "features/case-study-retold-remote.md"
+								}
+							]
 						}
 					],
 					"DocFiles": [
@@ -357,12 +382,15 @@
 						"features/beacon-providers.md",
 						"features/beacons.md",
 						"features/capabilities.md",
+						"features/case-study-retold-remote.md",
 						"features/cli.md",
 						"features/configuration.md",
 						"features/llm-model-setup.md",
 						"features/llm.md",
 						"features/manifests.md",
 						"features/operations.md",
+						"features/platform-cards.md",
+						"features/reachability-matrix.md",
 						"features/scheduling.md",
 						"features/tasks-content-system.md",
 						"features/tasks-data-transform.md",
@@ -374,6 +402,7 @@
 						"features/tasks-meadow-api.md",
 						"features/tasks-user-interaction.md",
 						"features/tasks.md",
+						"features/universal-addressing.md",
 						"index.html",
 						"overview.md",
 						"quickstart.md",