**Arguments:** $ARGUMENTS
Surfaces and tracks project assumptions with tiered confidence levels for alignment.
/plugin marketplace add Jeffallan/claude-skills/plugin install jeffallan-fullstack-dev-skills-2@Jeffallan/claude-skillscommon-ground/Arguments: $ARGUMENTS
Claude often operates on assumptions about project context, technology choices, coding standards, and user preferences. This command surfaces those assumptions for explicit user validation, preventing misaligned work based on incorrect premises.
Parse arguments to determine mode:
| Flag | Mode | Description |
|---|---|---|
| (none) | Default | Surface & Adjust two-phase interactive flow |
--list | List | Read-only view of all tracked assumptions |
--check | Check | Quick validation of current assumptions |
Load detailed guidance based on context:
| Topic | Reference | Load When |
|---|---|---|
| Assumption Types & Tiers | references/assumption-classification.md | Classifying assumptions, determining type or tier |
| File Management | references/file-management.md | Storage operations, project ID, ground file format |
Before any operation, determine the project identity:
Try git remote:
git remote get-url origin 2>/dev/null
If found, use the URL as project identifier (e.g., github.com/user/repo)
Fallback to path: Use the current working directory absolute path
Store this as {project_id} for file operations.
When no flags provided, execute the two-phase interactive flow.
Analyze current context to identify assumptions:
Classify each assumption by type and proposed tier:
references/assumption-classification.md for classification rulesPresent to user via AskUserQuestion:
Present assumptions grouped by category. Use multiSelect to let user choose which to track.
Question format:
I've identified assumptions about this project. Which should I track?
Options (up to 4 categories, user selects via multiSelect):
User can select multiple categories or use "Other" to specify individual items.
Handle uncertain items:
For any [uncertain] assumptions, ask direct clarifying questions.
Show selected assumptions with proposed tiers
Present tier adjustment options via AskUserQuestion:
Question format:
Review the confidence tiers. Any adjustments needed?
Options:
Process adjustments:
assumption text [tier] [type]Write ground file:
~/.claude/common-ground/{project_id}/ground.mdground.index.json for machine-readable accessreferences/file-management.md for file formats## Common Ground Complete
**Project:** {project_name}
**Tracked Assumptions:** {count}
### Summary
- ESTABLISHED: {count} (high confidence)
- WORKING: {count} (medium confidence)
- OPEN: {count} (needs validation)
**Ground file saved to:** ~/.claude/common-ground/{project_id}/ground.md
Run `/common-ground --list` to view all assumptions.
Run `/common-ground --check` for quick validation.
Read-only display of all tracked assumptions.
Load ground file from ~/.claude/common-ground/{project_id}/ground.md
Display assumptions grouped by tier:
## Common Ground: All Assumptions
**Project:** {project_name}
**Last Updated:** {timestamp}
### ESTABLISHED ({count})
1. {title} - {assumption} [{type}]
2. ...
### WORKING ({count})
1. {title} - {assumption} [{type}]
2. ...
### OPEN ({count})
1. {title} - {assumption} [{type}]
2. ...
---
(Read-only view. Run `/common-ground` to modify.)
No ground file found for this project.
Run `/common-ground` to surface and track assumptions.
Quick validation of existing assumptions.
Load ground file from ~/.claude/common-ground/{project_id}/ground.md
Present summary via AskUserQuestion:
Question format:
Quick check: Are these assumptions still valid?
Options:
Handle responses:
last_validated timestamp, confirmHandle missing file: If no ground file exists, redirect to default flow:
No existing assumptions to check. Starting fresh...
Then execute default mode.