Beginner-Friendly Topic
Take your time - it's perfectly normal to re-read this topic 2-3 times. Try the interactive code editor below to run code yourself. Use the Q&A section to check your understanding before moving on.You've got this!
Agent Goals
A goal defines what the agent is supposed to achieve. Goals must be clear, specific, and bounded to produce reliable agent behaviour.
6 min•By Priygop Team•Updated 2026
Goal Structure
Goal Structure
# Goal definition for a business research agent
goal = {
"primary_objective": "Research and summarise the top 3 competitors for our product",
"scope": "Only include companies with a similar product category",
"output_format": "Structured report with: company name, product, pricing, key features",
"constraints": [
"Only use publicly available information",
"Do not contact any company",
"Maximum 5 sources per competitor"
],
"approval_required": True, # Human must approve before the report is sent
"deadline": "Within 10 minutes"
}
# The agent reads this goal and uses it to guide every decision
print("Goal loaded. Agent will begin research.")Good and Poor Goal Design
- Poor goal: 'Find some information about competitors' - too vague, the agent does not know when to stop
- Good goal: 'Find the top 3 competitors for our project management tool and list their pricing plans and key features'
- Poor goal: 'Do everything to close this deal' - too broad, could lead to unsafe actions
- Good goal: 'Draft a follow-up email for this proposal and wait for approval before sending it'
- Key rule: a goal should describe the desired outcome clearly enough that the agent can recognise when it has been achieved
Key Takeaways
- A goal defines what the agent is supposed to achieve.
- Poor goal: 'Find some information about competitors' - too vague, the agent does not know when to stop
- Good goal: 'Find the top 3 competitors for our project management tool and list their pricing plans and key features'
- Poor goal: 'Do everything to close this deal' - too broad, could lead to unsafe actions