| name | ghpm-status |
| description | GitHub Project health dashboard. Shows workflow distribution, component health, team workload, items needing attention, and available views. |
| argument-hint | [scope (ex. team/component)] |
| allowed-tools | Bash(gh:*), Read, Grep |
| compatibility | Requires gh CLI authenticated with read:project and project scopes |
| metadata | {"author":"jackchuka","scope":"generic"} |
ghpm-status
PREREQUISITE: Read ../ghpm-shared/SKILL.md for prerequisites and error handling.
Comprehensive project health dashboard. The goal is to give the user a rich, actionable overview — not just counts, but breakdowns that surface where work is concentrated, who's overloaded, and what needs attention. Think of this as the dashboard a project manager checks every morning.
Arguments
/ghpm-status — full dashboard for the project/ghpm-status <scope> — scoped to a team or component name (matched against field options in .ghpm/config.json)
Workflow
Phase 1: Load Config and Cache
Follow the startup sequence in ../ghpm-shared/SKILL.md and load cache per ../ghpm-shared/references/cache.md.
Phase 2: Aggregate
Count items by workflow column (using the workflow.field name from config to find the matching field on each item). Include a count for items with no status.
If a scope argument was provided, filter items first:
- Match the argument against all field option names in
.ghpm/config.json (case-insensitive).
- If no exact match, try semantic/fuzzy matching. The user may use informal or umbrella terms that don't match any single field option but logically group several. Use your understanding of the project's team/component names to infer which options the user likely means, and combine them. Explain the mapping you chose (e.g., "Interpreting '' as + ").
- If you still can't determine a reasonable match, list available options and ask the user to clarify — but try hard to infer first.
- Filter items to only those matching the scope before aggregating.
Build the following sections from the cached items. Each section adds a different dimension of insight. Include all sections where the data exists — skip a section only if the project doesn't have that field.
Component Health
If a "Component" (or similar grouping) field exists, break down items by component × status. This reveals which areas of the project are healthy vs struggling. Show completion percentage per component and flag components with low completion or high active workload.
Epic Tracker
If items have a "Level" or type field that distinguishes epics from tasks, show epic-level status separately. Active epics (InProgress) are especially useful — list them with their component and assignees. Flag unassigned active epics as a risk.
Team Workload
Count InProgress items per assignee. This surfaces overload — anyone with 8+ concurrent InProgress items deserves a flag. Also count unassigned InProgress items.
Attention Items
Identify items needing attention:
- Untriaged: items with no status, no component, or no assigned team (check which fields exist in config — not all projects have these fields).
- No assignee: items in non-Done columns with empty assignees. Call out ReadyForDev items without owners specifically — they're bottlenecked.
- Active InProgress: items in InProgress status — count as a workload indicator. No date info is available, so don't claim staleness.
Available Views
List available views with their layout type and filter text. Do NOT attempt to count items per view — filter parsing is handled by /ghpm-view. Just show the view name, layout, and raw filter so the user knows what's available.
For key normalization when matching fields, see ../ghpm-shared/references/cache.md.
Phase 3: Format Output
Format per conventions in ../ghpm-shared/references/format.md. Use markdown tables for breakdowns — they're easier to scan than flat text. Include a brief "Key Observations" section at the end that highlights 3-5 actionable insights (bottlenecks, risks, overloaded contributors).
The output should feel comprehensive. A good dashboard is one where the user doesn't need to ask follow-up questions to understand the project's health.
If scoped, prefix with the scope name: <Project Title> — <scope> — <filtered count> of <total> items
End with: Use /ghpm-view <name> to drill into a view.
Tips
- Read-only — never modifies the project or its items.
See Also
