I Built Interactive CLI Drills — Learners Quit at Prompt #3

A reporter wanted to showcase interactive command line drills designed for learners. However, the analytics revealed that a significant percentage of learners abandoned the drills at the third prompt. The reporter aimed to examine the issue and explain the changes made to address it.

The reporter initially created short missions for CMD, PowerShell, and Bash-oriented tasks on Windows-friendly paths. Immediate feedback was provided upon completion, and a clean terminal UI was designed to mimic the practice experience rather than a MOOC video. The reporter believed that skill development occurs at the command line prompt.

The reporter noted that prompt #1 was successful, as learners could list files and print a string, resulting in dopamine release. Prompt #2 introduced one new concept and was still acceptable. However, prompt #3 combined concepts or required error recovery without a scaffold, causing confidence to plummet. The reporter realized that the difficulty curve was too aggressive, and learners were quitting.

To address the issue, the reporter implemented smaller concept steps, ensuring one idea per prompt until the streak felt earned. Later, these concepts were combined in a "mission boss" rather than at step three. Additionally, the reporter added intentional recovery prompts, introducing a safe failure that taught about ERRORLEVEL and $LASTEXITCODE after a success, rather than before.

The reporter also highlighted the importance of intentional recovery prompts and softer first-session definitions of done. A complete first session could consist of six tiny wins, with heroic achievements in the second session. The reporter emphasized that the product's purpose is to train operators, not to win a landing-page award.

The reporter acknowledged that the quit data did not destroy the thesis; instead, it revealed the need for pacing adjustments. By watching metrics like median time-to-first-success, drop-off by prompt index, repeat visits within 48 hours, and qualitative comments about feeling dumb, the reporter discovered that prompt #3 was the cliff in their curriculum graph. After re-pacing, the cliff moved later and became shorter.

Ultimately, the reporter concluded that prompt #3 was not a fatal flaw but an opportunity to re-evaluate pacing. They suggested that if a product requires complex combining of ideas and perfect quoting, it may not be a drill but a challenging experience. The reporter advised splitting the curriculum to make the third prompt a fair win and earning the harder combines.

They cautioned against adding gamification, such as badges and streaks, as a Band-Aid solution. Instead, the reporter recommended making the third prompt a fair win, then gradually increasing complexity.

Finally, the reporter shared a free, browser-based tool called CMD Master for interactive terminal learning in CMD, PowerShell, and Bash. They encouraged readers to watch where they want to quit, as that spot might be a future design ticket. The reporter posed a question to other trainers: do you gate progress strictly or allow learners to skip and return? They sought insights on balancing structured practice with adult learners' needs.