fix(summary): fix review model parsing

src/reviewbot/agent/tasks/issues.py

@@ -14,6 +14,7 @@
 from reviewbot.core.issues import IssueModel
 from reviewbot.core.issues.issue_model import IssueModelList
 from reviewbot.tools.diff import get_diff_from_file
+from reviewbot.tools.read_file import read_file_from_store
 
 console = Console()
 
@@ -371,6 +372,7 @@ async def quick_scan_file(
 
     try:
         diff_content = get_diff_from_file(agent.store, file_path)
+        file_content = read_file_from_store(agent.store, file_path)
     except Exception as e:
         console.print(f"[yellow]Could not fetch diff for {file_path}: {e}[/yellow]")
         return True  # If can't get diff, do deep review to be safe
@@ -399,11 +401,13 @@ async def quick_scan_file(
         HumanMessage(
             content=f"""Quickly scan this file and determine if it needs deep review: {file_path}
 
-Here is the diff:
 
-```diff
+Here is the file:
+{file_content}
+
+Here is the diff:
 {diff_content}
-```"""
+"""
         ),
     ]
 
@@ -482,32 +486,10 @@ async def review_single_file(
     reasoning_context = get_reasoning_context(agent.store)
 
     # Force a reasoning pass to ensure think() is invoked during deep review
-    try:
-        diff_content = get_diff_from_file(agent.store, file_path)
-        think_messages: list[BaseMessage] = [
-            SystemMessage(
-                content=(
-                    "You are a senior code reviewer. You must think and review. "
-                    "with 2-5 sentences of reasoning about the provided diff. "
-                    "Do not use any other tools. Once finished, reply with the "
-                    "single word DONE."
-                )
-            ),
-            HumanMessage(
-                content=f"""Diff for {file_path}:
-
-```diff
-{diff_content}
-```
-""",
-            ),
-        ]
-        if model:
-            think_settings = ToolCallerSettings(max_tool_calls=40)
-            ido_agent = create_ido_agent(model=model, tools=tools or [])
-            await ido_agent.with_tool_caller(think_settings).ainvoke(think_messages)
-    except Exception as e:
-        console.print(f"[yellow]Failed to record reasoning for {file_path}: {e}[/yellow]")
+    diff_content = get_diff_from_file(agent.store, file_path)
+    file_content = read_file_from_store(agent.store, file_path)
+    if not file_content:
+        file_content = ""
 
     messages: list[BaseMessage] = [
         SystemMessage(
@@ -584,15 +566,11 @@ async def review_single_file(
         HumanMessage(
             content=f"""Review the merge request diff for the file: {file_path}
 
-WORKFLOW:
-1. Use `get_diff("{file_path}")` to get the diff
-2. Analyze the changes for bugs, security issues, and logic errors
-3. If you need context beyond the diff (imports, variable declarations, surrounding code):
-   - Use `read_file("{file_path}")` to see the complete file
-4. Use `think()` to document your reasoning and analysis
-5. Return your findings as a list of issues (or empty list if none)
+File content:
+{file_content}
 
-Find the real bugs - that's what matters most!
+Diff:
+{diff_content}
 """
         ),
     ]
@@ -661,26 +639,19 @@ async def validate_issues_for_file(
     messages: list[BaseMessage] = [
         SystemMessage(
             content=(
-                "You are an issue validator. Your job is to remove FALSE POSITIVES while keeping real bugs.\n\n"
-                "AVAILABLE TOOLS:\n"
-                "- `read_file(file_path)` - Read the complete file to verify issues\n"
-                "- `ls_dir(dir_path)` - List directory contents to verify file structure\n\n"
-                "WHAT TO REMOVE (false positives):\n"
-                "- 'Variable X undefined' - when X is actually defined elsewhere in the file\n"
-                "- 'Import Y missing' - when Y exists at the top of the file\n"
-                "- 'Function Z not declared' - when Z is defined in the complete file\n\n"
-                "WHAT TO KEEP (real issues):\n"
-                "- Logic errors - wrong conditions, broken algorithms, incorrect business logic\n"
-                "- Security vulnerabilities - SQL injection, XSS, auth bypass, etc.\n"
-                "- Bugs that will crash or produce wrong results\n"
-                "- Data corruption risks\n"
-                "- Performance problems\n\n"
-                "RULES:\n"
-                "- KEEP issues about logic, bugs, security, and functionality\n"
-                "- ONLY remove issues that are provably false (use read_file to verify)\n"
-                "- When in doubt, KEEP the issue - don't filter out real bugs\n"
-                "- Do NOT create new issues\n"
-                "- Do NOT modify issue fields"
+                "You are a strict Issue Validator filter. The project is ALREADY COMPILED AND RUNNING.\n\n"
+                "### CRITICAL CONTEXT\n"
+                "1. THE CODEBASE HAS ALREADY COMPILED AND BUILT SUCCESSFULLY.\n"
+                "2. If a file is deleted in a diff, it means the references were already cleaned up.\n"
+                "3. PROVISION: Any issue claiming 'compilation error', 'missing reference', or 'broken startup' is FACTUALLY WRONG.\n\n"
+                "### VALIDATION ARCHITECTURE\n"
+                "Your ONLY goal is to discard issues that assume the code is currently broken. "
+                "Since the build passed, the code is structurally sound. You are only looking for LOGIC bugs in NEW code.\n\n"
+                "### DISCARD IMMEDIATELY (False Positives):\n"
+                "- **Deletions:** Claims that deleting code/files will break the app (The build passed, so it didn't).\n"
+                "- **References:** Claims that a symbol is undefined (It is likely defined in a file you can't see).\n"
+                "- **Build/Runtime:** Any mention of 'compilation errors', 'build failures', or 'initialization failures'.\n"
+                "- **Assumptions:** Speculation about files outside the provided diff.\n\n"
             )
         ),
         HumanMessage(
@@ -695,17 +666,17 @@ async def validate_issues_for_file(
 {json.dumps(issues_payload, indent=2)}
 
 TASK:
-1. For issues about "undefined/missing" code, use `read_file("{file_path}")` to check if the code actually exists elsewhere
-2. Remove ONLY clear false positives
-3. Keep all logic bugs, security issues, and real functionality problems
+1. If the diff shows a file being DELETED, and the issue claims this deletion causes a failure elsewhere: DISCARD THE ISSUE.
+2. The fact that the file was deleted and the project STILL COMPILED means the initialization logic was moved or is no longer necessary.
+3. Validate only the logic within the lines starting with '+' (added).
 
 Return a ValidationResult with:
 - valid_issues: confirmed real issues
 - removed: false positives with reason for removal"""
         ),
     ]
 
-    validation_settings = ToolCallerSettings(max_tool_calls=5)  # Allow tool calls for validation
+    validation_settings = ToolCallerSettings(max_tool_calls=10)  # Allow tool calls for validation
 
     try:
         if model is None:

src/reviewbot/agent/workflow/diff_extract.py

@@ -3,46 +3,6 @@
 from typing import Any
 
 
-def _extract_code_from_diff(diff_text: str, line_start: int, line_end: int) -> str:
-    hunk_header_pattern = re.compile(r"^@@\s+-(\d+)(?:,(\d+))?\s+\+(\d+)(?:,(\d+))?\s+@@")
-    lines = diff_text.splitlines()
-
-    extracted = []
-    current_new = 0
-    in_hunk = False
-
-    for line in lines:
-        match = hunk_header_pattern.match(line)
-        if match:
-            current_new = int(match.group(3))
-            in_hunk = True
-            continue
-
-        if not in_hunk:
-            continue
-
-        # We only care about the lines in the NEW file (the result of the change)
-        if line.startswith("+"):
-            if line_start <= current_new <= line_end:
-                extracted.append(line[1:])  # Remove '+'
-            current_new += 1
-        elif line.startswith("-"):
-            # Skip deleted lines for code extraction of the 'new' state
-            continue
-        else:
-            # Context line
-            if line_start <= current_new <= line_end:
-                extracted.append(line[1:] if line else "")
-            current_new += 1
-
-        # FIX: Exit early if we've passed the end of our requested range
-        if current_new > line_end:
-            if extracted:  # Only break if we actually found lines
-                break
-
-    return "\n".join(extracted)
-
-
 def generate_line_code(file_path: str, old_line: int | None, new_line: int | None) -> str:
     """
     Generates a GitLab-compatible line_code.

src/reviewbot/agent/workflow/gitlab_notes.py

@@ -1,6 +1,7 @@
 from typing import Any, NamedTuple
 from urllib.parse import quote
 
+from ido_agents.agents.tool_runner import ToolCallerSettings
 from langchain_core.language_models.chat_models import BaseChatModel
 from langgraph.func import task  # type: ignore
 from rich.console import Console  # type: ignore
@@ -28,7 +29,7 @@ class AcknowledgmentResult(NamedTuple):
 
 
 @task
-def post_review_acknowledgment(
+async def post_review_acknowledgment(
     *, gitlab: GitProviderConfig, diffs: list[FileDiff], model: BaseChatModel
 ) -> AcknowledgmentResult | None:
     """
@@ -147,11 +148,10 @@ def post_review_acknowledgment(
     try:
         # Get response with no tool calls allowed
         from ido_agents.agents.ido_agent import create_ido_agent
-        from ido_agents.agents.tool_runner import ToolCallerSettings
 
         summary_settings = ToolCallerSettings(max_tool_calls=0)
         ido_agent = create_ido_agent(model=model, tools=[])
-        summary = ido_agent.with_tool_caller(summary_settings).invoke(messages)
+        summary = await ido_agent.with_tool_caller(summary_settings).ainvoke(messages)
 
         # Post as a discussion (so we can update it later)
         acknowledgment_body = f"""<img src="https://img.shields.io/badge/Code_Review-Starting-blue?style=flat-square" />
@@ -186,7 +186,8 @@ def post_review_acknowledgment(
         return None
 
 
-def update_review_summary(
+@task
+async def update_review_summary(
     api_v4: str,
     token: str,
     project_id: str,
@@ -234,6 +235,43 @@ def update_review_summary(
     total_files = len(diffs)
     files_with_issues = len(issues_by_file)
 
+    # Build change overview for context
+    new_files: list[str] = []
+    deleted_files: list[str] = []
+    renamed_files: list[tuple[str, str]] = []
+    modified_files: list[str] = []
+
+    for diff in diffs:
+        if diff.is_renamed:
+            renamed_files.append((diff.old_path or "unknown", diff.new_path or "unknown"))
+        elif diff.is_new_file:
+            new_files.append(diff.new_path or diff.old_path or "unknown")
+        elif diff.is_deleted_file:
+            deleted_files.append(diff.old_path or diff.new_path or "unknown")
+        else:
+            modified_files.append(diff.new_path or diff.old_path or "unknown")
+
+    change_stats = (
+        "Files changed: "
+        f"{total_files} (new: {len(new_files)}, modified: {len(modified_files)}, "
+        f"renamed: {len(renamed_files)}, deleted: {len(deleted_files)})"
+    )
+    change_overview_lines: list[str] = []
+    change_overview_lines.extend(f"- {path} (new)" for path in new_files)
+    change_overview_lines.extend(f"- {path} (deleted)" for path in deleted_files)
+    change_overview_lines.extend(f"- {old} -> {new} (renamed)" for old, new in renamed_files)
+    change_overview_lines.extend(f"- {path} (modified)" for path in modified_files)
+
+    max_change_lines = 12
+    if len(change_overview_lines) > max_change_lines:
+        remaining = len(change_overview_lines) - max_change_lines
+        change_overview_lines = change_overview_lines[:max_change_lines]
+        change_overview_lines.append(f"- ... and {remaining} more file(s)")
+
+    change_overview_text = (
+        "\n".join(change_overview_lines) if change_overview_lines else "- No files listed."
+    )
+
     # Prepare issue details for LLM
     issues_summary: list[str] = []
     for issue in issues:
@@ -256,7 +294,7 @@ def update_review_summary(
 - Provide reasoning about the overall merge request purpose and code quality.
 - Highlight key concerns or positive aspects
 - Be constructive and professional
-- DO NOT use any tools
+- Use tools to generate a comprehensive summary
 - Use paragraphs with readable flow. Use two paragrahs with 3-5 sentences.
 Paragraphs should be wrapped with <p> tags. Use new <p> tag for a newline.
 Example
@@ -267,7 +305,8 @@ def update_review_summary(
 <p>
 paragraph2
 </p>
-- Focus on the big picture, not individual issue details"""
+- Focus on the big picture, not individual issue details
+- Reference the changes overview so the summary stays grounded in what changed, even if there are no issues"""
             ),
             HumanMessage(
                 content=f"""A code review has been completed with the following results:
@@ -280,24 +319,35 @@ def update_review_summary(
   - Medium severity: {medium_count}
   - Low severity: {low_count}
 
+**Changes overview:**
+{change_stats}
+{change_overview_text}
+
 **Issues found:**
 {issues_text}
 
 - Use EXACTLY two paragraphs, each wrapped in <p> tags.
 1. Provides overall assessment of the purpose of the merge request purpose and code quality.
 2. Highlights the most important concerns (if any)
 3. Gives reasoning about the review findings
-4. Is constructive and actionable      """
+4. Is constructive and actionable
+5. Mention the kinds of changes and at least one example file from the changes overview      """
             ),
         ]
 
         if model is None:
             raise ValueError("model parameter is required for ido-agents migration")
 
         ido_agent = create_ido_agent(model=model, tools=tools or [])
-        llm_summary = ido_agent.with_structured_output(ReviewSummary).invoke(messages)
 
-        llm_summary = str(llm_summary)
+        summary_settings = ToolCallerSettings(max_tool_calls=5)
+        llm_summary = await (
+            ido_agent.with_structured_output(ReviewSummary)
+            .with_tool_caller(summary_settings)
+            .ainvoke(messages)
+        )
+
+        llm_summary = llm_summary.summary if llm_summary else "Review completed successfully."
 
     except Exception as e:
         console.print(f"[yellow]Failed to generate LLM summary: {e}[/yellow]")

src/reviewbot/core/reviews/review_model.py

@@ -1,9 +1,5 @@
-from pydantic import RootModel
+from pydantic import BaseModel
 
 
-class ReviewSummary(RootModel[str]):
-    def __str__(self) -> str:
-        return self.root
-
-    def __len__(self) -> int:
-        return len(self.root)
+class ReviewSummary(BaseModel):
+    summary: str