Merge pull request #111 from TaskarCenterAtUW/feat/akb/4094-akb-helpline-answers

Feat/akb/4094 akb helpline answers

Author: uw-amy-bordenave

CHANGELOG.md

@@ -7,6 +7,7 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and this project adheres to [Semantic Versioning](https://semver.org/) and [Conventional Commits](https://www.conventionalcommits.org/).
 
 <!--
+
 ## Unreleased
 
 ### Features
@@ -21,6 +22,16 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 
 -->
 
+## v14.3.0 (2026-07-24)
+
+### Features
+
+- **Docs**: Update Assistant Knowledge Base articles
+
+### Fixes
+
+- **Core**: Fix AKB dispatch generator to use CRLF line endings
+
 ## v14.2.0 (2026-07-22)
 
 ### Features

docs/assistant/accessmap/concept/accessmap.md

@@ -1,5 +1,5 @@
 ---
-title: "What is AccessMap?"
+title: What is AccessMap?
 slug: accessmap
 doc_type: concept
 questions:
@@ -15,9 +15,9 @@ topics:
     - accessmap
     - overview
 risk_level: medium
-authority_level: provisional
+authority_level: explanatory
 publication_status: draft
-last_reviewed: 2026-07-22
+last_reviewed: 2026-07-24
 retrieval_priority: high
 assistant_behavior:
     allow_inference: false
@@ -39,31 +39,31 @@ tags:
 
 ## Short Answer
 
-AccessMap is a free, web-based pedestrian trip planner developed by TCAT. It supports accessibility-aware multimodal routing using pedestrian network information such as paths, crossings, curb ramps, barriers, and elevation.
+AccessMap is a free, web-based pedestrian trip planner developed by TCAT. It supports accessibility-aware multimodal routing using pedestrian network information such as sidewalks, crossings, curb ramps, barriers, and elevation.
 
 ## Significance
 
 AccessMap lets travelers compare route options using mobility profiles and preferences rather than treating accessibility as one fixed condition. It also supports screen-reader-oriented directions and user feedback about mapped features and route steps.
 
 ## What This Means
 
-Users can select a region, choose or customize a mobility profile, set waypoints, review route and elevation information, follow directions with device location, and submit feedback. The application uses OpenSidewalks and OpenStreetMap data.
+Users can select a region, choose or customize a mobility profile, set waypoints, review route and elevation information, preview or follow real-time directions with device location, and submit feedback. The application uses OpenSidewalks and OpenStreetMap data.
 
 ## What This Does Not Mean
 
-Coverage and feature accuracy depend on the available regional data. A route result is informational and does not certify physical accessibility or replace judgment about current conditions.
+Coverage and feature accuracy depend on the available regional data. A route result is informational and does not legally certify physical accessibility or replace judgment about current conditions.
 
 ## How To Use This
 
 Use AccessMap to plan and compare pedestrian routes, identify mapped constraints, and communicate observations. Check the selected region, profile, and route details before relying on a result.
 
 ## Example
 
-A traveler selects a wheelchair profile, sets an origin and destination, reviews steep sections and route statistics, and changes preferences if the result does not fit their needs.
+A traveler selects a wheelchair profile, sets an origin and destination, reviews steep sections and route statistics, and changes preferences if the result does not fit their individual needs.
 
 ## Assistant Guidance
 
-State the region and profile when explaining a result. Do not infer that an unmapped barrier is absent, and recommend checking conditions directly when safety or critical access is involved.
+State the region and profile when explaining a result. Do not infer that an unmapped feature or barrier is absent. Recommend using best judgment around current conditions.
 
 ## Related Concepts
 

docs/assistant/accessmap/concept/longer-route-selection.md

@@ -4,6 +4,7 @@ slug: longer-route-selection
 doc_type: concept
 questions:
     - Why does AccessMap choose longer routes?
+    - Why might AccessMap choose a longer route?
 audiences:
     - public
     - advocate
@@ -15,11 +16,11 @@ topics:
 risk_level: low
 authority_level: explanatory
 publication_status: draft
-last_reviewed: 2026-06-19
+last_reviewed: 2026-07-24
 retrieval_priority: high
 assistant_behavior:
     allow_inference: true
-    requires_citation: true
+    requires_citation: false
     abstain_if_missing_context: false
     do_not_claim:
         - The shortest car route is always the correct comparison baseline.
@@ -40,28 +41,28 @@ AccessMap trades raw shortest distance for lower modeled cost under your mobilit
 
 ## Significance
 
-Preempts support tickets and social posts framed as routing "bugs."
+Routes generated by AccessMap are often longer than car-focused routes provided by other routers. This is intended behavior, not an error or bug.
 
 ## What This Means
 
-The optimizer minimizes a cost function, not only meters. Costs come from network attributes and profile settings documented in the user manual.
+The optimizer minimizes a cost function, not only distance. Costs come from network attributes and mobility profile settings.
 
 ## What This Does Not Mean
 
-- Not proof that the shorter path is illegal or impassable for everyone.
-- Not an error solely because another map shows a different line.
+- Not proof that a shorter path is illegal or impassable for everyone.
+- Not an error solely because another map returns a different route between the same destinations.
 
 ## How To Use This
 
-Suggest checking active profile, preferences, and elevation-related settings before reporting data issues.
+Suggest that the user check the active mobility profile, preferences, and elevation-related settings before suggesting reporting data issues.
 
 ## Example
 
-Two routes differ by two blocks, but the shorter one crosses a segment tagged with a high slope penalty for manual wheelchair profiles.
+Two routes differ by two blocks, but the shorter one crosses a segment tagged with a high slope penalty for manual wheelchair profiles. This is intended behavior, as the longer route is the correct route based on the inputs.
 
 ## Assistant Guidance
 
-Explain multi-criteria optimization in plain language. Avoid telling users their preferred route is "wrong"; explain tradeoffs instead.
+Explain multi-criteria optimization in plain language. Avoid telling users their preferred route is "wrong"; explain tradeoffs instead, and that AccessMap supports customizable routing that individuals can tailor to their needs.
 
 ## Related Concepts
 

docs/assistant/cross-platform/concept/abbreviations.md

@@ -47,7 +47,7 @@ topics:
 risk_level: low
 authority_level: official
 publication_status: published
-last_reviewed: 2026-07-22
+last_reviewed: 2026-07-24
 retrieval_priority: high
 assistant_behavior:
     allow_inference: false
@@ -68,7 +68,7 @@ tags:
 
 ## Short Answer
 
-This page lists all 272 abbreviations, acronyms, and initialisms used across TCAT Wiki documentation and TCAT platforms and tools.
+This page lists all 273 abbreviations, acronyms, and initialisms used across TCAT Wiki documentation and TCAT platforms and tools.
 
 ## Significance
 
@@ -269,6 +269,7 @@ The table below maps each abbreviated form to its full expansion. Entries are so
 | RSO | Registered Student Organization |
 | RTAN | Regional Transportation Accessibility Network |
 | RTD | Denver Regional Transportation District |
+| RTOR | Right Turn on Red |
 | RTPO | Regional Transportation Planning Organization |
 | SAFETEA-LU | Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users |
 | SAM | Segment Anything Model |

docs/assistant/cross-platform/concept/connected-pedestrian-graph.md

@@ -4,30 +4,35 @@ slug: connected-pedestrian-graph
 doc_type: concept
 questions:
     - What is a connected pedestrian graph?
+    - What does "connectedness" mean?
 audiences:
     - planner
     - jurisdiction
+    - developers
     - advocate
 products:
+    - OpenSidewalks
     - OS-CONNECT
     - AccessMap
     - Walksheds
 topics:
     - graph-metrics
+    - opensidewalks
     - os-connect
     - accessmap
     - walksheds
 risk_level: low
 authority_level: explanatory
 publication_status: draft
-last_reviewed: 2026-06-16
+last_reviewed: 2026-07-24
 retrieval_priority: high
 assistant_behavior:
     allow_inference: true
-    requires_citation: true
+    requires_citation: false
     abstain_if_missing_context: false
     do_not_claim:
-        - The graph matches every informal desire line or off-street private path.
+        - Connectedness means that paths are physically connected in real life.
+        - Spatially overlapping features are necessarily connected in the data.
 related_pages:
     - assistant/cross-platform/concept/accessibility-islands.md
     - assistant/cross-platform/concept/walksheds.md
@@ -41,34 +46,39 @@ tags:
 
 ## Short Answer
 
-The connected pedestrian graph is the set of walkable edges and nodes — sidewalks, crossings, ramps, and transitions — that routing and walkshed engines use as a network. Connectivity is a property of that graph under specific rules and data, not of aerial imagery alone.
+Connectedness is a property of a network graph: each element in the data is connected to the graph so routing can proceed across the network. It is about connections as represented in the data, not about whether two paths in real life can be traversed between.
+
+An unconnected dataset contains features that overlap spatially but are disconnected: they do not share connections in the data.
 
 ## Significance
 
-Many public questions about "missing routes" or "wrong walksheds" are graph questions. Explaining the graph shifts conversations from app bugs alone to data stewardship and design standards.
+Routing and walkshed results depend on data connections, not geometry alone. This distinction helps teams diagnose network-quality problems accurately.
 
 ## What This Means
 
-Edges carry attributes (slope, surface, crossing type, etc.) that affect whether they are traversable for a profile or cost model. The same geometry can be connected for one profile and effectively disconnected for another.
+- Network elements must share represented connections in the data for routing to move between them.
+- Spatial overlap alone does not create a graph connection.
+- A disconnected dataset can prevent routes from connecting between features that appear to meet on a map.
 
 ## What This Does Not Mean
 
-- Not equivalent to every socially used walking path.
-- Not a promise that the fastest car road connection is mirrored for pedestrians.
+- It does not mean that paths which appear to meet on a map are connected in the dataset.
+- It does not mean that overlapping geometries without connections in the data are routable.
 
 ## How To Use This
 
-When scoping improvements, trace graph components and critical connectors. When communicating with the public, distinguish "not routable in the app" from "physically impossible to walk."
+When reviewing a dataset, inspect the data for shared connections and disconnected components rather than relying only on visual overlap.
 
 ## Example
 
-A campus path exists on the ground but is not in OS-CONNECT; the routing graph may omit it until mapped and validated through the jurisdiction's pipeline.
+In an unconnected dataset, where two sidewalks that visually overlap at a street corner but do not have a shared node connecting them in the data, a router is unable to create a route that crosses from one sidewalk to the other, even though the features may appear to meet when rendered on a basemap.
 
 ## Assistant Guidance
 
-If users report a missing connection, acknowledge graph dependence and point to feedback or data contribution pathways in the relevant product docs without promising timelines.
+Explain that connectedness in the pedestrian network data context refers to a property of the data (that features which overlap spatially share common data elements that enable routing between them) not a property of the real world.
 
 ## Related Concepts
 
-- [Accessibility islands](accessibility-islands.md)
+- [Network topology](../../opensidewalks/concept/network-topology.md)
+- [Completeness](completeness.md)
 - [Walksheds](walksheds.md)

docs/assistant/cross-platform/concept/os-connect-vs-gtfs-pathways.md

@@ -0,0 +1,92 @@
+---
+title: How is OS-CONNECT different from GTFS-Pathways?
+slug: os-connect-vs-gtfs-pathways
+doc_type: concept
+questions:
+    - How is OS-CONNECT different from GTFS-Pathways?
+    - Should I use OS-CONNECT or GTFS-Pathways for transit access data?
+audiences:
+    - planner
+    - jurisdiction
+    - advocate
+    - public
+    - developer
+products:
+    - OS-CONNECT
+    - AccessMap
+topics:
+    - cross-platform
+    - os-connect
+    - gtfs
+    - accessmap
+    - accessibility-data
+risk_level: medium
+authority_level: explanatory
+publication_status: draft
+last_reviewed: 2026-07-24
+retrieval_priority: high
+assistant_behavior:
+    allow_inference: true
+    requires_citation: false
+    abstain_if_missing_context: false
+    do_not_claim:
+        - OS-CONNECT and GTFS-Pathways describe the same geographic scope.
+        - OS-CONNECT automatically includes a transit agency's station-internal pathways.
+        - GTFS-Pathways is a complete connected pedestrian network for Washington State.
+related_pages:
+    - assistant/cross-platform/index.md
+    - assistant/os-connect/index.md
+    - assistant/accessmap/concept/gtfs-pathways.md
+    - assistant/support/workflow/choose-dataset-for-transit-stations.md
+tags:
+    - Assistant
+---
+
+<!-- @format -->
+
+# How is OS-CONNECT different from GTFS-Pathways?
+
+## Short Answer
+
+OS-CONNECT is Washington's connected pedestrian network dataset. GTFS-Pathways is an expansion of the General Transit Feed Specification (GTFS) for describing pathways and transfers inside a transit station or stop.
+
+Use OS-CONNECT for pedestrian-network context outside the station and use a transit agency's GTFS-Pathways data for station-internal connections when the published feed provides those records. A trip-planning analysis may need both datasets because they describe different parts of the journey.
+
+## Significance
+
+The distinction affects dataset selection, routing, and the interpretation of accessibility information. Treating one source as a replacement for the other can omit either the surrounding pedestrian network or the station's internal layout.
+
+## What This Means
+
+- OS-CONNECT represents a connected pedestrian network in Washington State, including sidewalks, crossings, curb ramps, and related pedestrian infrastructure.
+- GTFS-Pathways represents connections among transit-facility nodes, such as entrances, platforms, elevators, stairs, escalators, and walkways, through GTFS records such as `pathways.txt`, `levels.txt`, and related `stops.txt` records.
+- OS-CONNECT and GTFS-Pathways can be combined at a station entrance, but the connection between datasets must be checked rather than assumed.
+- Actual coverage depends on the published dataset or feed version. Inspect the records and documentation before relying on a feature or accessibility value.
+
+## What This Does Not Mean
+
+- OS-CONNECT is not automatically the authoritative source for a transit agency's station-internal pathways.
+- GTFS-Pathways is not a complete connected pedestrian network for Washington State or the streets around a station.
+- A matching map location does not prove that the two datasets share a routable connection.
+- An accessibility attribute in either source does not guarantee that current physical conditions are accessible or unchanged.
+
+## How To Use This
+
+- **Planners and jurisdictions:** Define the geographic boundary of the analysis first. Use GTFS-Pathways for station-internal movement and OS-CONNECT for the surrounding pedestrian approach, then document the feed and dataset versions.
+- **Trip-planning integrators and developers:** Inspect the agency feed for `pathways.txt`, `levels.txt`, and the related `stops.txt` hierarchy. Check station entrances and nearby pedestrian features before joining the two networks.
+- **Advocates and public users:** Ask which source supports a specific accessibility claim and whether the source is current. For a current physical condition, verify with the transit agency or another qualified source.
+
+## Example
+
+A regional planner evaluates an accessible route from a sidewalk to a station platform. The planner uses OS-CONNECT to examine the sidewalk and crossing leading to the station entrance, then checks the transit agency's GTFS-Pathways records for the entrance, elevator, and platform connections. If either source lacks the needed records, the planner reports the gap instead of assuming that the route is continuous.
+
+## Assistant Guidance
+
+Explain the difference in geographic scope. Verify the current OS-CONNECT release and transit feed contents before answering any specific questions about coverage or accessibility. If the user does not identify a dataset version, station, or jurisdiction when that context affects the answer, ask for it or state the limitation. Do not present either dataset as a legal compliance determination or a guarantee of current physical accessibility.
+
+## Related Concepts
+
+- [Cross-Platform knowledge base](../index.md)
+- [OS-CONNECT knowledge base](../../os-connect/index.md)
+- [What is GTFS-Pathways?](../../accessmap/concept/gtfs-pathways.md)
+- [Which dataset should I use if I need transit-station pathway data?](../../support/workflow/choose-dataset-for-transit-stations.md)