Which task suits a coding agent best as a first step?
Why C is the answer
Adding tests for a well-understood module, reviewed in a pull request
Start where the scope is clear, the result can be checked and a mistake is easy to reverse.
In detail
- A well-understood module has low risk.
- A pull request adds review, so it is a good first step.
- Choose work that is easy to verify.
Use case
An illustrative scenario, not a real incident.
A team starts with an agent adding tests to a well-known module. Risk is low, the output is easy to review, and trust grows from there.
See it in different ways
flowchart TD
A{"How risky is the action?"} -->|"Low and reversible"| B["Agent acts, people review later"]
A -->|"Medium"| C["Agent proposes in a pull request"]
A -->|"High"| D["Human approval before acting"]sequenceDiagram
participant A as Agent
participant V as Validation
participant R as Repository
A->>V: Structured plan
alt Plan passes
V->>A: Go ahead
A->>R: Makes changes in a branch
else Plan fails
V-->>A: Fix the plan first
endflowchart TD
Q{"What is the task like?"}
Q -->|"Well understood and reviewable"| A0["Good first task"]
Q -->|"Live secrets or no tests"| A1["Not a first task"]Key point: Start with low-risk, reviewable work.
Why the other options are wrong
- A. Rewriting the payment system with no tests in one change
- No tests and one huge change is high risk.
- B. Choosing the company's long-term architecture on its own
- Long-term architecture is a human decision.
- D. Rotating production secrets directly in the live environment
- Live secrets are too risky for a first step.