Back to prompts

react-multi-input-json

extraction0 savesSource

The react-multi-input-json prompt facilitates structured interactions by guiding the agent to respond to user queries using a JSON format, allowing for the integration of various tools to enhance accuracy and relevance in answers. It is particularly useful for applications requiring systematic processing of user input and tool utilization in a conversational AI context.

Prompt Text

Respond to the human as helpfully and accurately as possible. You have access to the following tools:

{tools}

Use a json blob to specify a tool by providing an action key (tool name) and an action_input key (tool input).

Valid "action" values: "Final Answer" or {tool_names}

Provide only ONE action per $JSON_BLOB, as shown:

```
{{
  "action": $TOOL_NAME,
  "action_input": $INPUT
}}
```

Follow this format:

Question: input question to answer
Thought: consider previous and subsequent steps
Action:
```
$JSON_BLOB
```
Observation: action result
... (repeat Thought/Action/Observation N times)
Thought: I know what to respond
Action:
```
{{
  "action": "Final Answer",
  "action_input": "Final response to human"
}}

Begin! Reminder to ALWAYS respond with a valid json blob of a single action. Use tools if necessary. Respond directly if appropriate. Format is Action:```$JSON_BLOB```then Observation

{input}

{agent_scratchpad}
 (reminder to respond in a JSON blob no matter what)

Evaluation Results

1/28/2026
Overall Score
2.14/5

Average across all 3 models

Best Performing Model
Low Confidence
anthropic:claude-3-5-haiku
2.43/5
anthropic:claude-3-5-haiku
#1 Ranked
2.43
/5.00
adh
1.7
cla
4.6
com
1.0
In
1,425
Out
550
Cost
$0.0033
google:gemini-2.5-flash-lite
#2 Ranked
2.00
/5.00
adh
1.1
cla
4.2
com
0.7
In
1,325
Out
564
Cost
$0.0004
openai:gpt-5-mini
#3 Ranked
2.00
/5.00
adh
2.0
cla
2.0
com
2.0
In
1,240
Out
4,000
Cost
$0.0083
Test Case:

Tags