User guide

This demo organizes development activity and satisfaction data with the SPACE framework so trends can be reviewed from multiple perspectives.

Basic workflow

1Select a project, branch, and date range on the dashboard.
2Review the source metrics and each SPACE dimension, not only the overall score.
3Use Comparison with aligned subjects and periods to investigate changes.
4Treat AI evaluation as supporting information and verify it against actual MRs and team context.

Data sources

In addition to anonymized mock data, the public demo can connect a GitHub repository from Settings. GitHub API requests use a server-managed token. Satisfaction responses and product feedback are stored in the demo PostgreSQL database.

Reading scores

Metrics are converted to a 0–100 scale and combined with configured weights. Missing values are excluded instead of being treated as zero, and the remaining weights are redistributed.

AI evaluation

AI mode combines analysis of demo MRs or GitHub pull requests with SPACE metrics to generate strengths, concerns, and improvement suggestions through the configured Gemini API.

Interpretation

Commit or changed-line counts alone do not establish outcome or quality. Consider role, period, difficulty, leave, and other context. AI output can also be inaccurate.

SPACE metrics

Performance

Evaluates the outcomes and quality produced by development work. Merged changes, developer-caused defects, and bug-fix speed indicate whether valuable results are delivered reliably.

Merged MRs Authored: Number of merge requests merged in the selected period. For a developer, this is MRs created by that developer and merged.
Assigned Bug Issues: Number of bug issues attributed to this project or developer in the selected period.
Bug Fix Lead Time (Hours): Average hours from bug issue creation to completion for fixed bugs in the selected period.

Activity

Evaluates the volume and continuity of development work through commits, issues, bug handling, and code changes. Activity is a proxy for work performed, not an outcome or quality measure by itself.

Commits: Total commits in the selected period, filtered by project, branch, and developer when selected.
Issues Created: Number of issues created in the selected period.
Bug Issues Reported: Number of bug issues found or reported in the selected period.
Bug Issues Fixed: Number of bug issues completed or fixed in the selected period.
Lines Added: Total added lines from commits in the selected period.
Lines Deleted: Total deleted lines from commits in the selected period.
Lines Changed: Total changed lines from commits in the selected period. This is additions plus deletions.

Communication

Evaluates collaboration and knowledge sharing through review participation, comments on others' merge requests, and constructive discussion during development.

Review Participation: Review participation count. For a developer, this counts other people's MRs where they left an effective review comment in the selected period. For a project, this counts non-author reviewer participation on project MRs.
Review Comments Given: Number of effective review comments posted on other people's MRs in the selected period. Author replies on their own MRs are not included.
MR Discussion Density: Discussion density: average number of non-author review comments received per created MR. This is shown as context and is not directly used for scoring, because higher is not always better.

Efficiency

Evaluates whether work flows smoothly with limited delays and interruptions. Merge lead time and review wait time show how efficiently changes reach completion.

Merge Lead Time (Hours): Average hours from MR creation to merge for MRs merged in the selected period. This uses the same MR set as merge count.
Review Wait Time: Average hours from MR creation to the first review comment.
Uninterrupted Focus Time: Average uninterrupted focus time per workday. Higher values indicate more time available for sustained deep work.
Context Switch Frequency: Average number of switches between tasks, projects, or meetings per workday. Lower values indicate less fragmented work.

Satisfaction

Evaluates developer satisfaction and well-being through job meaning, developer efficacy, sustainability, and confidence that improvements can happen.

Job Satisfaction & Meaning: Sub-score calculated from work value, meaningful work, team satisfaction, and recommendation intent.
Developer Efficacy: Sub-score calculated from information access, development environment support, support flow, and confidence in outcomes.
Sustainability: Sub-score calculated from sustainable workload, fatigue, detachment, and excessive pressure. Negative questions are reverse-scored.
Improvement Potential: Sub-score calculated from psychological safety and expectation that improvement suggestions can lead to real change.