From ffdacdc2a4b1e6c318060c1be6e4041bb940eacd Mon Sep 17 00:00:00 2001 From: James Borlase Date: Wed, 29 Jul 2026 16:06:07 +0200 Subject: [PATCH 1/3] Cross-link agentic contexts from personalize-travel agent tutorial The demo travel app injects Signals attributes into prompts and does not support agentic contexts yet. Add a note pointing readers to the agentic contexts docs and the Vercel AI SDK tutorial for the native pattern, keeping the attribute-based flow as the user-level profile complement. Co-Authored-By: Claude Fable 5 --- tutorials/signals-personalize-travel/personalizing-agent.md | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/tutorials/signals-personalize-travel/personalizing-agent.md b/tutorials/signals-personalize-travel/personalizing-agent.md index f795dc6e5..52753bbd3 100644 --- a/tutorials/signals-personalize-travel/personalizing-agent.md +++ b/tutorials/signals-personalize-travel/personalizing-agent.md @@ -4,7 +4,7 @@ sidebar_label: "Personalize the AI agent" position: 6 description: "Integrate Snowplow Signals with an AI chatbot to provide personalized travel recommendations based on user behavioral data." keywords: ["chatbot", "AI personalization", "OpenAI", "conversational AI", "behavioral data"] -date: "2025-01-21" +date: "2026-07-29" --- ```mdx-code-block @@ -21,6 +21,10 @@ When a user asks the agent a question, the system fetches the attribute values f A toggle button on the chat widget allows you to enable or disable the tool call so you can see the difference in responses with and without Signals personalization. +:::note[Agentic contexts] +Signals can also build a prompt block like this for you. An [agentic context](/docs/signals/agentic-contexts/) captures a user's recent in-session activity, and you can [retrieve it](/docs/signals/applications/agentic-contexts/) as an LLM-ready narrative. The demo app doesn't support agentic contexts yet, so this tutorial uses attributes only. To see the native pattern, follow the [AI agent with real-time user context tutorial](/tutorials/signals-ai-agent-context/introduction). The two approaches complement each other: attributes give the agent a computed profile of the user, while an agentic context grounds it in the current session. +::: + ## Data flow Date: Fri, 31 Jul 2026 11:41:06 +0100 Subject: [PATCH 2/3] Make the agentic contexts cross-link one sentence of prose Replace the note panel with a single sentence at the end of How it works, pointing at the standalone agentic contexts tutorial rather than the Vercel one, which doesn't cover agentic contexts. Drop the roadmap phrasing, the concept re-explanation, and "LLM-ready" in favour of the wording on the page being linked, and restore the creation date. Co-Authored-By: Claude Fable 5 --- tutorials/signals-personalize-travel/personalizing-agent.md | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/tutorials/signals-personalize-travel/personalizing-agent.md b/tutorials/signals-personalize-travel/personalizing-agent.md index 52753bbd3..4b1f733d1 100644 --- a/tutorials/signals-personalize-travel/personalizing-agent.md +++ b/tutorials/signals-personalize-travel/personalizing-agent.md @@ -4,7 +4,7 @@ sidebar_label: "Personalize the AI agent" position: 6 description: "Integrate Snowplow Signals with an AI chatbot to provide personalized travel recommendations based on user behavioral data." keywords: ["chatbot", "AI personalization", "OpenAI", "conversational AI", "behavioral data"] -date: "2026-07-29" +date: "2025-01-21" --- ```mdx-code-block @@ -21,9 +21,7 @@ When a user asks the agent a question, the system fetches the attribute values f A toggle button on the chat widget allows you to enable or disable the tool call so you can see the difference in responses with and without Signals personalization. -:::note[Agentic contexts] -Signals can also build a prompt block like this for you. An [agentic context](/docs/signals/agentic-contexts/) captures a user's recent in-session activity, and you can [retrieve it](/docs/signals/applications/agentic-contexts/) as an LLM-ready narrative. The demo app doesn't support agentic contexts yet, so this tutorial uses attributes only. To see the native pattern, follow the [AI agent with real-time user context tutorial](/tutorials/signals-ai-agent-context/introduction). The two approaches complement each other: attributes give the agent a computed profile of the user, while an agentic context grounds it in the current session. -::: +Signals can also assemble this personalization context for you: an [agentic context](/docs/signals/agentic-contexts/) captures a user's recent in-session activity, which you can [retrieve](/docs/signals/applications/agentic-contexts/) as a plain-language narrative — see [Give an AI agent real-time session context with Signals](/tutorials/signals-agentic-contexts/start). This tutorial uses attributes. ## Data flow From de743001ee93d3011de4a521c9cc2c9f1e3dd857 Mon Sep 17 00:00:00 2001 From: James Borlase Date: Fri, 31 Jul 2026 11:56:03 +0100 Subject: [PATCH 3/3] Drop cross-branch link to the unmerged agentic contexts tutorial The docs links cover the concept; add the tutorial link once its PR merges. Co-Authored-By: Claude Fable 5 --- tutorials/signals-personalize-travel/personalizing-agent.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tutorials/signals-personalize-travel/personalizing-agent.md b/tutorials/signals-personalize-travel/personalizing-agent.md index 4b1f733d1..79192a374 100644 --- a/tutorials/signals-personalize-travel/personalizing-agent.md +++ b/tutorials/signals-personalize-travel/personalizing-agent.md @@ -21,7 +21,7 @@ When a user asks the agent a question, the system fetches the attribute values f A toggle button on the chat widget allows you to enable or disable the tool call so you can see the difference in responses with and without Signals personalization. -Signals can also assemble this personalization context for you: an [agentic context](/docs/signals/agentic-contexts/) captures a user's recent in-session activity, which you can [retrieve](/docs/signals/applications/agentic-contexts/) as a plain-language narrative — see [Give an AI agent real-time session context with Signals](/tutorials/signals-agentic-contexts/start). This tutorial uses attributes. +Signals can also assemble this personalization context for you: an [agentic context](/docs/signals/agentic-contexts/) captures a user's recent in-session activity, which you can [retrieve](/docs/signals/applications/agentic-contexts/) as a plain-language narrative. This tutorial uses attributes. ## Data flow