Evidence beats adjectives
'Passionate developer with expertise in modern technologies' says nothing. A repository with a real README, a working demo link and a written explanation of why the architecture is what it is says a great deal. Reviewers — human or otherwise — trust artifacts they can inspect.
Honest labels matter as much as the work itself. Marking a project as a concept or an exploration is not a weakness; it is a signal that your other claims can be trusted.
Example scenario: two project pages
Thought experiment: two candidates list a CLI tool. One page offers a screenshot and three adjectives; the other links to a repository and explains the problem, trade-off and limitations. The second page provides more inspectable information, not a guarantee of an interview.
This is an illustrative scenario, not a claim about specific hiring outcomes.
Depth versus breadth
A smaller set of inspectable projects can communicate your approach more clearly than many shallow cards. Neither count nor depth alone determines hiring outcomes. The aim is to let reviewers verify scope, decisions and limitations through the linked work.
A short portfolio checklist
Every project links to something inspectable; every claim is verifiable or removed; concept work is labelled as concept; each write-up names a real trade-off and a real limitation; and broken links and stale screenshots are treated as bugs, not decoration.
Source: EnvDoctor case study
Source: ChronoFlow case study