fix: optimize skill

Author: craigyu

.agents/skills/pr-comment-resolver/SKILL.md

@@ -22,13 +22,39 @@ This skill outlines the step-by-step workflow for fetching, analyzing, addressin
 * Call the MCP tool `list_pull_requests` on `github-mcp-server` (or query the GitHub API) to locate the open pull request that corresponds to the active branch. Note down the `pullNumber`.
 
 ### 2. Retrieve Review Comments
+**Option A: Using GitHub CLI (Recommended)**
+```bash
+# Get all pull request review comments
+gh api repos/bcgov/nr-silva/pulls/<PR_NUMBER>/comments
+
+# Parse with Python for readability
+gh api repos/bcgov/nr-silva/pulls/<PR_NUMBER>/comments 2>&1 | python3 -c "
+import json, sys
+for c in json.load(sys.stdin):
+    print(f'ID: {c[\"id\"]} | Path: {c[\"path\"]} | Line: {c[\"line\"]} | Body: {c[\"body\"][:80]}...')
+"
+```
+
+**Option B: Using MCP Tools**
 * Call the `pull_request_read` tool with `method="get_review_comments"`, specifying the `owner`, `repo`, and `pullNumber`.
-* Identify all unresolved comment threads (`is_resolved: false`). For each thread, extract:
-  * The `commentId` (obtained from the comment object or the suffix of `html_url` e.g., `#discussion_r3599506357` -> `3599506357`).
+
+**For each comment, extract:**
+  * The `commentId` (the numeric ID, e.g., `3678193517`).
   * The file `path` and target `line` numbers.
   * The comment `body` detailing the requested change.
+  * The `is_resolved` flag — only address unresolved comments (`is_resolved: false`).
+
+### 3. Validate Comments Retrieved
+**Before proceeding, confirm:**
+```bash
+# Count total unresolved comments
+gh api repos/bcgov/nr-silva/pulls/<PR_NUMBER>/comments | \
+  python3 -c "import json, sys; comments = json.load(sys.stdin); print(f'Total: {len(comments)} comment(s)')"
+```
+* If count is **0**, the PR may not have review comments yet, or they may already be resolved.
+* If count is **> 0**, proceed with validation and addressing each comment.
 
-### 3. Validate Each Comment
+### 4. Validate Each Comment
 For each comment:
 1. **Read Context**: Examine the referenced code, the full method/component, and surrounding logic.
 2. **Assess Validity**: Determine if the comment is:
@@ -41,7 +67,7 @@ For each comment:
    - If **valid**: Proceed to Step 4 (Address the Comment).
    - If **invalid/unnecessary**: Reply with a brief explanation of why the comment is not applicable, then move to next comment.
 
-### 4. Address Each Valid Comment Sequentially
+### 5. Address Each Valid Comment Sequentially
 For each validated comment:
 1. **Analyze and Modify**: Read the referenced code block and edit the file to address the feedback using your code editing tools.
 2. **Local Validation**: Run the relevant local tests or verification commands to confirm the fix is correct and functional.
@@ -54,13 +80,25 @@ For each validated comment:
    ```bash
    git push
    ```
-5. **Get Commit Hash**: Extract the 7-digit commit SHA (found in the commit command output or via `git rev-parse --short HEAD`).
-6. **Post Reply**: Call the MCP tool `add_reply_to_pull_request_comment` with the arguments:
+5. **Get Commit Hash**: Extract the 7-digit commit SHA:
+   ```bash
+   git rev-parse --short HEAD
+   ```
+6. **Post Reply** (two options):
+
+   **Option A: GitHub CLI (Recommended)**
+   ```bash
+   gh api repos/bcgov/nr-silva/pulls/<PR_NUMBER>/comments/<COMMENT_ID>/replies \
+     -f body="Fixed in commit <7-digit-hash>"
+   ```
+
+   **Option B: MCP Tool**
+   Call the `add_reply_to_pull_request_comment` tool with:
    * `owner`, `repo`, `pullNumber`
    * `commentId`: the numeric ID of the review comment
    * `body`: `"Fixed in commit <7-digit-hash>."`
 
-### 5. Final Verification
+### 6. Final Verification
 * Verify the codebase is clean and regression-free:
   * **Check number of changed files**: Count the files modified across all commits.
   * **If ≤2 files changed**: Run targeted tests for the affected modules only.
@@ -92,3 +130,34 @@ Alternatively, to verify compilation only (faster):
 cd backend
 ./mvnw clean compile
 ```
+
+---
+
+## Troubleshooting
+
+### Problem: No comments found when calling the API
+**Common causes:**
+1. **Wrong PR number**: Verify the PR number from GitHub or use `gh pr view` to confirm the current PR
+   ```bash
+   gh pr view  # Shows current PR number and branch
+   ```
+2. **Comments already resolved**: Check GitHub UI to see if all Copilot/reviewer comments are marked as resolved
+3. **API permissions**: Ensure GitHub CLI is authenticated:
+   ```bash
+   gh auth status
+   ```
+
+### Problem: Forgot to fetch comments initially
+**Solution:** Always run the fetch and validation step (Step 2-3) FIRST, before analyzing any code:
+```bash
+# Quick check if PR has any review comments
+gh api repos/bcgov/nr-silva/pulls/1380/comments | python3 -c "import json, sys; print(len(json.load(sys.stdin)), 'comment(s)')"
+```
+This prevents missed reviews and wasted time analyzing code without knowing what needs fixing.
+
+### Problem: Getting 404 on API call
+**Typical issue:** Using wrong repository path. Verify with:
+```bash
+gh repo view  # Shows owner/repo path
+```
+Then use the correct format: `repos/{owner}/{repo}/pulls/{number}/comments`

backend/src/main/java/ca/bc/gov/restapi/results/common/service/impl/AbstractOpeningSearchService.java

@@ -247,8 +247,7 @@ private void validatePageSize(Pageable pagination) {
     int pageSize = pagination.getPageSize();
     if (pageSize <= 0) {
       throw new ResponseStatusException(
-          HttpStatus.BAD_REQUEST,
-          "Page size must be greater than 0, but was: " + pageSize);
+          HttpStatus.BAD_REQUEST, "Page size must be greater than 0, but was: " + pageSize);
     }
     if (pageSize > SilvaConstants.MAX_PAGE_SIZE_OPENING_SEARCH) {
       throw new MaxPageSizeException(SilvaConstants.MAX_PAGE_SIZE_OPENING_SEARCH);