GPT-OSS model behavior and instruction guidelines

Edit online

When you select the GPT-OSS model for your agent, the selected model influences how the agent behaves and interprets instructions. GPT-OSS 120B — OpenAI (via Groq) and GPT-OSS 120B — OpenAI (AWS Bedrock) models are highly capable but also sensitive to vague or conflicting instructions. To ensure reliable behavior:

  • Be explicit: Clearly state priorities and constraints.

  • Avoid ambiguity: Conflicting directives can lead to unpredictable outputs.

  • Test iteratively: Validate instructions with sample tasks before deploying.

  • Limit complexity: Break down multi-step reasoning into clear, sequential guidance.

Example for GPT-OSS 120B — OpenAI (via Groq) and GPT-OSS 120B — OpenAI (AWS Bedrock) models:

Always respond in a professional tone. When asked for CRM data,
use the salesforce_api tool. If the request involves scheduling,
call the calendar_agent. If information is incomplete, ask
clarifying questions before proceeding.
Important:

The GPT-OSS 120B — OpenAI model is available through GroqCloud and AWS Bedrock. When using GroqCloud, the model is governed by a third-party license that might impose use restrictions and other obligations. By using this model, you agree to the terms. Read the terms

Special considerations for GPT-OSS

The GPT-OSS model does not use the standard system prompt from watsonx Orchestrate. As a result, it can behave differently from other models, including:

  • Not explicitly identifying itself as part of the watsonx Orchestrate ecosystem.

  • Preferring its internal knowledge over your connected knowledge bases, unless you are instructed otherwise.

  • Hyperlinks might render incorrectly unless you specify formatting rules.

  • Agent styles selected in the UI do not apply when you use the GPT-OSS model.

To guide GPT-OSS reliably, add explicit natural-language constraints in the Instructions field. Add one or more of the following blocks to your agent’s instructions:

Prioritize knowledge bases over internal knowledge

Use this when you want the model to rely on your enterprise content rather than its internal pretrained knowledge.

Always check the connected knowledge base(s) first.
Prefer information retrieved from knowledge over your
own internal knowledge. If relevant content is found,
summarize it faithfully and cite the source title or document
name. If the knowledge base does not contain the answer, say
"I don't know based on the provided knowledge" and ask
a clarifying question.

Cap reasoning depth and iteration count

Use this when you want predictable performance and response times without long or looping reasoning chains.

Use concise reasoning. Limit yourself to
a maximum of 3 reasoning steps before answering.
Do not re-plan unless the last tool result contradicts
prior assumptions. If you cannot progress after 3 steps,
ask one focused clarifying question.

Formatting

Use this when the output needs a consistent hyperlink format.

When generating hyperlinks, use the correct Markdown syntax: [link](url).

Keep chit-chat short

Use this when you want minimal conversational filler so the model focuses on the task.

Keep small talk to a single sentence.
Immediately pivot to the user's task with a
concise question or action.

Enforce a strict output budget

Use this when you want clear, concise answers.

Target an answer length of 4-6 sentences (or ≤150 words).
Use bullet points only when they increase clarity.
Avoid repeating the prompt or restating obvious context.

Fail fast when data is missing

Use this when a task cannot proceed without inputs and you want quick recovery instead of speculation.

If required inputs are missing, do not speculate.
Ask for the minimum missing fields in a single
question and wait.

Combined instruction example

Use this compact template when creating an agent with GPT-OSS 120B — OpenAI (via Groq) and GPT-OSS 120B — OpenAI (AWS Bedrock). You can paste it into the Instructions field and customize it for your use case.

Behavior and sources:
- Identify as an agent operating within watsonx Orchestrate.
- Always use available tools to retrieve information before relying on your internal knowledge.
- If tools don't contain the answer, explicitly state "The available tools don't contain the answer", state that clearly and ask one clarifying question.

Reasoning and brevity controls:
- Use concise reasoning with at most 3 reasoning steps before responding.
- Keep chit-chat to one sentence, then proceed.
- Limit final answers to ≤150 words unless the user requests detail.

Formatting:
- When generating hyperlinks, use the correct Markdown syntax: [link](url).
 
Error handling:
- If a tool call or retrieval fails, briefly describe the failure and propose the next best step or ask for a missing input.

Good versus bad instruction examples for GPT-OSS model

Good (concise, enforced constraints):

Use knowledge bases as the primary source.
Cite the document name when used.
Limit yourself to 3 reasoning steps
and less than 150 words in the final answer.
If you cannot find an answer in knowledge,
say so and ask one clarifying question.

Bad (encourages runaway reasoning and verbosity):

Think in exhaustive detail until you're
absolutely certain. Explore all possible interpretations.
Provide a comprehensive narrative of your
thoughts and include context for every claim.