Master the terminology behind Sourcegraph's agentic coding tool for large codebases.
0 / 5 completed
1 / 5
At standup, a dev mentions using an agentic coding tool from Sourcegraph designed for handling large, complex coding tasks with minimal hand-holding. Which tool fits?
Amp is Sourcegraph's agentic coding tool built to autonomously tackle substantial coding tasks, leveraging code search and context-gathering to work effectively across large codebases. It emphasizes handling complex, multi-file work with less manual guidance. This distinguishes it from lightweight autocomplete-style assistants.
2 / 5
During a design review, the team wants the agent to search across a large monorepo to find relevant usages before making a change. Which capability supports this?
Amp leverages codebase-wide search to locate relevant files, usages, and patterns before proposing a change, which matters in large monorepos where relevant context can be scattered. This grounds the agent's proposed edits in the actual surrounding code. Strong retrieval is especially important at monorepo scale.
3 / 5
In a code review, a dev wants to review the reasoning and steps Amp took before accepting a large multi-file change. Which feature supports this?
Amp surfaces a trace of the steps and reasoning behind a multi-file change, letting a reviewer understand why particular files were touched and how the agent reached its conclusion. This transparency supports trust when accepting large, agent-produced diffs. It follows the same principle as other agentic tools that expose their working process.
4 / 5
An incident report shows an agent-produced change touched files outside the intended scope of the task. What practice reduces this risk?
Narrowly scoping the task given to an agent and reviewing the complete diff before merging catches unintended out-of-scope changes before they ship. Broad, vague instructions increase the chance the agent touches more than intended. This discipline applies across essentially all agentic coding tools, not just Amp.
5 / 5
During a PR review, a teammate asks how Amp's approach differs from a simple inline code-completion plugin. What is the distinction?
Unlike inline completion that predicts the next few tokens as you type, Amp is designed to plan and carry out multi-step tasks spanning several files, using search and reasoning to get there. This agentic scope is the key differentiator. It shifts the developer's role toward directing and reviewing rather than typing every change.