Phase 1b: Rename all ~94 commands across 12 plugins to /<noun> <action> sub-command pattern. Git-flow consolidated from 8→5 commands (commit variants absorbed into --push/--merge/--sync flags). Dispatch files, name: frontmatter, and cross-reference updates for all plugins. Phase 2: Design documents for 8 new plugins in docs/designs/. Phase 3: Scaffold 8 new plugins — saas-api-platform, saas-db-migrate, saas-react-platform, saas-test-pilot, data-seed, ops-release-manager, ops-deploy-pipeline, debug-mcp. Each with plugin.json, commands, agents, skills, README, and claude-md-integration. Marketplace grows from 12→20. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
52 lines
1.1 KiB
Markdown
52 lines
1.1 KiB
Markdown
---
|
|
name: pr findings
|
|
description: List and filter review findings by category/severity
|
|
agent: coordinator
|
|
---
|
|
|
|
# /pr findings - Filter Review Findings
|
|
|
|
## Visual Output
|
|
|
|
Display header: `PR-REVIEW - Findings`
|
|
|
|
## Skills to Load
|
|
|
|
- skills/mcp-tools-reference.md
|
|
- skills/review-patterns/confidence-scoring.md
|
|
- skills/output-formats.md
|
|
|
|
## Usage
|
|
|
|
```
|
|
/pr findings <pr-number> [--category <cat>] [--severity <sev>] [--confidence <min>] [--file <pattern>] [--compact] [--json]
|
|
```
|
|
|
|
## Workflow
|
|
|
|
### Without Previous Review
|
|
|
|
Prompt: "No review found. Run `/pr review`, `/pr summary`, or cancel?"
|
|
|
|
### With Previous Review
|
|
|
|
1. Load findings from previous review
|
|
2. Apply filters (category, severity, confidence, file)
|
|
3. Display using format from `skills/output-formats.md`
|
|
|
|
## Output Formats
|
|
|
|
Reference `skills/output-formats.md`:
|
|
- **Detailed** (default) - Full finding
|
|
- **Compact** (`--compact`) - Single line
|
|
- **JSON** (`--json`) - Structured data
|
|
|
|
## Examples
|
|
|
|
```bash
|
|
/pr findings 123 --category security
|
|
/pr findings 123 --severity critical,major
|
|
/pr findings 123 --confidence 0.8
|
|
/pr findings 123 --file src/api/*
|
|
```
|