Improve response extraction

agent/prompts/normalization.txt

@@ -16,4 +16,7 @@ Relevent examples are included in preceeding messages, use these as exact inspir
 The claim to normalize is:
 ###TITLE###
 
-Produce no other text other than the condensed claim in inverted commas. Nothing should be around the normalised claim.
+Produce no other text other than the condensed claim, surrounded <norm></norm>
+
+For example: BREAKING: the sky is green!
+Becomes: <norm>The sky is green</norm>

agent/prompts/trigger.txt

@@ -17,10 +17,15 @@ Include a url to a source for your trigger event (not a web search, a specific u
 Include the date that the event happened ("March 2022" for exmaple)
 
 Use a JSON format with each entry containing "Event,ReasoningWhyRelevant,SearchQuery,Url,Date".
-Return ONLY valid JSON.
-Do not include explanations.
-Do not wrap in markdown.
-Do not label where the json is using colons
+Return ONLY JSON, no extra text. Wrap it like this:
+<json>
+[
+    {
+        "Event": "Example"
+        ...
+    }
+]
+</json>
 
 Multiple tool invocations should be requested at once, if applicable.
 Use your abilities to look between the lines and produce some insightful analysis, thinking both short and long term.
@@ -32,4 +37,6 @@ Analysis should only be completed for proposed events that would graner >0.7 poi
 
 Since URLs change frequently, use tools to retreive up to date informaiton everytime, provided examples or existing knowledge will be wrong or out of date.
 
+Remember to return just json enclosed by <json></json>
+
 Lets go through it step by step