Project Status
Run a multi-agent workflow for project analysis and reporting.
Quick Start
bash
# Full workflow (scan + analyze + report + commit) uv run ~/.config/opencode/skill/util-project-status/scripts/orchestrator.py # Dry run (no file writes) uv run ~/.config/opencode/skill/util-project-status/scripts/orchestrator.py --dry-run # Skip git commit uv run ~/.config/opencode/skill/util-project-status/scripts/orchestrator.py --no-commit
Scan Only
bash
# Just scan projects without full analysis uv run ~/.config/opencode/skill/util-project-status/scripts/scan_projects.py # Show last scan result uv run ~/.config/opencode/skill/util-project-status/scripts/scan_projects.py --latest # Machine-readable output uv run ~/.config/opencode/skill/util-project-status/scripts/scan_projects.py --json
Workflow Steps
| Step | Agent | Description |
|---|---|---|
| 0 | L1 | Run scan_projects.py → discovery.json |
| 1 | L1×9 | Analyze projects in parallel → analyses.json |
| 2 | L3 | Write STATUT.md using template |
| 3 | L2 | Append post-mortem |
| 4 | L2 | Commit files to git |
| 5 | - | Summary to user |
Output Structure
code
EXPORT/statut_de_projet/
├── STATUT_<run_id>/
│ ├── discovery.json
│ └── STATUT.md
└── historique/
└── STATUT_<previous_run>/
Constraints
- •Max 9 parallel agents
- •60s timeout per agent
References
- •Analysis prompt: See references/ANALYSIS_PROMPT.md
- •Report template: See references/TEMPLATE.md