Untitled Lesson
Skill Matrix
UNLOCK NODES BY LEARNING NEW TAGS.
Why does an agentic coding tool generally perform better on a task like "make the 3 failing tests in this file pass" than on a vague request like "improve this file"?
💻 Code Challenge | +75 XP
Write a project context file (as commented content) for an agentic coding tool, including test/lint commands to run, key architectural conventions, and a rule about never committing directly to main.
An agentic tool was given a vague "improve the error handling" task and produced changes that seemed plausible but didn't address the actual reported bug. Reorder the steps to get a more useful, verifiable outcome.
Task: Reorder the blocks in logical sequence to solve the problem.
A.D.A. Interface
Adaptive Didactic Assistant

Pascual Vila
Frontend Instructor // Code Syllabus
The Error //
Giving an agentic coding tool a vague, open-ended task with no concrete way to verify success
// Vague: no way to verify success
"Improve the error handling in this file"
// Verifiable: a concrete, checkable outcome
"Make error responses include a specific error code field,
verified by the existing error-format tests passing"The Solution //
Without a checkable success criterion, the agent has no way to verify its own work through iteration, and is more likely to produce a plausible-looking but ultimately unhelpful or incorrect change. Frame tasks with a specific, verifiable outcome the agent can check by actually running something.
The Error //
Reviewing each file in a multi-file agentic change in isolation, rather than as one coherent unit of work
// Insufficient: reviewing routes/orders.ts alone, missing context
// Correct: reviewing the full multi-file diff as ONE coherent change,
// understanding how the route, service, and tests relate to each otherThe Solution //
An agentic tool's changes often span several related files as one coherent piece of reasoning — a route, its service, and its tests, for example — and reviewing each file separately can miss how they fit together or whether the overall change actually makes sense as a whole.