danielscholl OSDU Engagement Skill
Measure engineering involvement across OSDU platform projects using the osdu-engagement CLI.
For installation, authentication, output formats, and common troubleshooting, see the shared CLI reference.
Quick Start
Before first use, verify the tool is available:
osdu-engagement --version
If the command is not found, stop and use the setup skill to install missing dependencies.
Do NOT attempt to install tools yourself — the setup skill handles installation with the correct
sources, user approval, and verification.
If installed, skip exploration and go straight to the intent detection table below.
When to Use This Skill
- Code contributions or commit activity
- Top contributors or active developers
- Review patterns or reviewer workload
- Architecture Decision Record (ADR) engagement
- Team participation metrics
- Contribution trends over time
Do NOT use for:
- Test reliability or flaky tests → use
osdu-quality - Open MRs, pipeline status, or issues → use
osdu-activity - Single-repo GitLab operations → use
glabdirectly
Intent Detection
| User Query | Command |
|---|---|
| "contributions", "commits", "who's active" | osdu-engagement contribution --output markdown |
| "top contributors", "rankings" | osdu-engagement contribution --output markdown |
| "trends", "monthly activity" | osdu-engagement contribution trend --output markdown |
| "ADR", "architecture decisions" | osdu-engagement decision --output markdown |
| "reviewers", "review activity" | osdu-engagement contribution --output markdown (includes review data) |
Commands
| Command | Purpose | Key Filters |
|---|---|---|
osdu-engagement contribution |
Code contribution analysis | --days, --project, --start-date/--end-date |
osdu-engagement contribution trend |
Historical trend analysis | --months |
osdu-engagement decision |
ADR engagement analysis | --days, --project |
For full CLI options and JSON output structures, see reference/commands.md.
Output Handling
Always pass --output markdown for AI consumption. It produces token-optimized output
that's easy to summarize. Use --output json only when you need raw data for field extraction.
Never omit the flag — the default is tty which outputs ANSI codes that break parsing.
Interpreting Results
Healthy Patterns:
- Multiple active contributors (not single-person dominance)
- Review load distributed across team (not one bottleneck reviewer)
- Consistent or growing activity trend
- ADRs getting engagement and discussion
Warning Signs:
- Single contributor dominance — bus factor risk
- Review bottleneck — few people doing all reviews
- Declining trend — team shrinking or disengaging
- ADRs with zero engagement — architecture decisions made without input
Workflow Examples
"Who's contributing most to partition?"
osdu-engagement contribution --project partition --output markdown
Highlight top contributors, their focus areas (MRs vs reviews vs comments), and whether contribution is concentrated or distributed.
"How engaged is the team overall?"
osdu-engagement contribution --days 30 --output markdown
osdu-engagement contribution trend --months 6 --output markdown
Combine current snapshot with trend to show whether engagement is growing, stable, or declining.
"Are ADRs getting attention?"
osdu-engagement decision --days 90 --output markdown
Surface ADRs with low engagement (proposed but no discussion), stale ADRs, and which projects are actively using the ADR process.
Cross-Domain Queries
For comprehensive team health, combine with sibling skills:
# Full engagement picture
osdu-engagement contribution --project partition --output markdown # who's contributing
osdu-quality analyze --project partition --output markdown # test health
osdu-activity mr --project partition --output markdown # open work
Synthesize: "Partition has 5 active contributors, tests are healthy at 95%, and 3 MRs are pending review."
Reference Documentation
- reference/commands.md — Full CLI options, JSON output structures
- reference/troubleshooting.md — Engagement-specific issues
- ../../reference/osdu-cli-reference.md — Installation, auth, output formats, common errors