aggregate-logs
Generate LEARNINGS.md from skill execution logs over a configurable time window.
Detailed step-by-step guide for fix-pr command execution.
> /plugin marketplace add athola/claude-night-marketHow it fires
How this command gets triggered: by you, by Claude, or both.
/workflow-stepsContext preview
What this command does when you run it.
Detailed step-by-step guide for fix-pr command execution.
Detailed step-by-step guide for fix-pr command execution.
> **See Also**: [Main Command](../fix-pr.md) | [Configuration](configuration-options.md) | [Troubleshooting](troubleshooting-fixes.md)
The fix-pr workflow consists of 6 sequential steps. Each step builds on the previous and can be skipped when appropriate.
| Step | Name | Purpose | When to Skip | |------|------|---------|--------------| | **[1](steps/1-analyze.md)** | [Analyze](steps/1-analyze.md) | Understand PR and gather comments | Already familiar with PR | | **[2](steps/2-triage.md)** | [Triage](steps/2-triage.md) | Classify by type and priority | Single simple fix | | **[3](steps/3-plan.md)** | [Plan](steps/3-plan.md) | Generate fix strategies | Fixes are obvious | | **[4](steps/4-fix.md)** | [Fix](steps/4-fix.md) | Apply code changes | Made changes manually | | **[5](steps/5-validate.md)** | [Validate](steps/5-validate.md) | Test and verify fixes | Already validated | | **[6](steps/6-complete.md)** | [Complete](steps/6-complete.md) | **Reply to & resolve threads** | **⛔ NEVER SKIP** |
> **⚠️ Step 6 is MANDATORY.** The workflow is incomplete until all review threads are replied to and resolved. See [Mandatory Exit Gate](../fix-pr.md#mandatory-exit-gate).
┌─────────────┐ ┌─────────────┐ ┌─────────────┐
│ 1. Analyze │ ──▶ │ 2. Triage │ ──▶ │ 3. Plan │
│ (Context) │ │ (Classify) │ │ (Strategy) │
└─────────────┘ └─────────────┘ └─────────────┘
│
▼
┌─────────────┐ ┌─────────────┐ ┌─────────────┐
│ 6. Complete │ ◀── │ 5. Validate │ ◀── │ 4. Fix │
│ (Threads) │ │ (Test) │ │ (Changes) │
└─────────────┘ └─────────────┘ └─────────────┘Understand the PR and gather all review comments.
**Key Actions:**
**Output:** PR metadata, all review comments, initial analysis
---
Classify comments by type and priority.
**Categories:**
**Output:** Classified comment list with priorities
---
Generate fix strategies for actionable comments.
**Commit Strategies:**
**Output:** Fix plan with strategies per comment
---
Apply code changes systematically.
**Process:** 1. Read code context (±20 lines) 2. Apply fix with Edit tool 3. Verify no new issues 4. Commit changes
**Output:** Applied fixes, commits created
---
Ensure all fixes are correct and quality gates pass.
**Validation Steps:**
headless browser (CDP MCP / Playwright); see Section 5.4
**Output:** Tests passing, validate-pr summary table ready for Gate 3
---
Resolve threads, create issues, and post summary.
**MANDATORY for PR authors receiving review comments.**
**Sub-Modules:** | Sub-Step | Module | Purpose | |----------|--------|---------| | **6.0** | [Reconciliation](steps/6-complete/reconciliation.md) | Reconcile ALL unworked items and enforcement | | **6.1-6.2** | [Issue Creation](steps/6-complete/issue-creation.md) | Create issues for suggestions/deferred items | | **6.3** | [Thread Resolution](steps/6-complete/thread-resolution.md) | Reply to and resolve ALL review threads | | **6.4** | [Issue Linkage](steps/6-complete/issue-linkage.md) | Link/close related issues | | **6.5** | [Summary](steps/6-complete/summary.md) | Post summary comment to PR | | **6.6** | [Verification](steps/6-complete/verification.md) | Final verification and workflow gate |
**Output:** All threads resolved, ALL unworked items tracked as issues, summary posted
---
**MANDATORY** - You MUST reply to and resolve each review thread:
# Reply to thread
gh api graphql -f query='
mutation {
addPullRequestReviewThreadReply(input: {
pullRequestReviewThreadId: "PRRT_xxx"
body: "Fixed - [description]"
}) { comment { id } }
}'
# Resolve thread
gh api graphql -f query='
mutation {
resolveReviewThread(input: {threadId: "PRRT_xxx"}) {
thread { isResolved }
}
}'See [Step 6: Complete](steps/6-complete.md) for full details.
If version issues were flagged, re-validate before proceeding:
| Project Type | Files to Check | |--------------|----------------| | claude-marketplace | marketplace.json, plugin.json files | | python | pyproject.toml, __version__ in code | | node | package.json, package-lock.json | | rust | Cargo.toml, Cargo.lock |
See [Step 5: Validate](steps/5-validate.md) for validation scripts.
---
This workflow documentation is modularized for efficient context loading:
fix-pr-modules/
├── workflow-steps.md # This file (hub/navigation)
├── configuration-options.md # Command configuration
├── troubleshooting-fixes.md # Common issues
└── steps/
├── 1-analyze.md # DiscovA plugin marketplace for Claude Code. Install only the plugins you need to run git workflows, code review, spec-driven development, and autonomous agents from inside your Claude Code session.
Generate LEARNINGS.md from skill execution logs over a configurable time window.
Analyze skill file complexity metrics and generate modularization recommendations for splitting or progressive loading.
Scaffold new Claude Code skills with brainstorming, TDD methodology, and proper frontmatter and module structure.