chore: import upstream snapshot with attribution

contributing/samples/adk_team/adk_issue_monitoring_agent/PROMPT_INSTRUCTION.txt

@@ -0,0 +1,18 @@
+You are the automated security and moderation agent for the {OWNER}/{REPO} repository.
+
+You will be provided with an Issue Number and a list of comments made by non-maintainers.
+Your job is to read through these comments and identify if any of them contain SPAM, promotional content for 3rd-party websites, SEO links, or objectionable material.
+
+CRITERIA FOR SPAM:
+- The comment is completely unrelated to the repository or the specific issue.
+- The comment promotes a 3rd party product, service, or website.
+- The comment is generic "SEO spam" (e.g., "Great post! Check out my site at [link]").
+
+INSTRUCTIONS:
+1. Evaluate the provided comments.
+2. If you identify spam, call the `flag_issue_as_spam` tool.
+   - Pass the `item_number`.
+   - Pass a brief `detection_reason` explaining which comment is spam and why (e.g., "@spammer_bot posted an irrelevant link to a shoe store").
+3. If NONE of the comments contain spam, do NOT call any tools. Just respond with "No spam detected."
+
+Remember: Do not flag comments that are merely unhelpful, off-topic, or from beginners asking legitimate questions. Only flag actual spam, endorsements, or objectionable material.

contributing/samples/adk_team/adk_issue_monitoring_agent/README.md

@@ -0,0 +1,65 @@
+# ADK Issue Monitoring Agent 🛡️
+
+An intelligent, cost-optimized, automated moderation agent built with the **Google Agent Development Kit (ADK)**.
+
+This agent automatically audits GitHub repository issues to detect SEO spam, unsolicited promotional links, and irrelevant third-party endorsements. If spam is detected, it automatically applies a `spam` label and alerts the repository maintainers.
+
+## ✨ Key Features & Optimizations
+
+- **Zero-Waste LLM Invocations:** Fetches issue comments via REST APIs and pre-filters them in Python. It automatically ignores comments from maintainers, `[bot]` accounts, and the official `adk-bot`. The Gemini LLM is never invoked for safe threads, saving 100% of the token cost.
+- **Dual-Mode Scanning:** Can perform a **Deep Clean** (auditing the entire history of all open issues) or a **Daily Sweep** (only fetching issues updated within the last 24 hours).
+- **Token Truncation:** Uses Regular Expressions to strip out Markdown code blocks (```` ``` ````) replacing them with `[CODE BLOCK REMOVED]`, and truncates unusually long text to 1,500 characters before sending it to the AI.
+- **Idempotency (Anti-Double-Posting):** The bot reads the comment history for its own signature. If it has already flagged an issue, it instantly skips it, preventing infinite feedback loops.
+
+______________________________________________________________________
+
+## Configuration
+
+The agent is configured via environment variables, typically set as secrets in GitHub Actions.
+
+### Required Secrets
+
+| Secret Name      | Description                                                                                          |
+| :--------------- | :--------------------------------------------------------------------------------------------------- |
+| `GITHUB_TOKEN`   | A GitHub Personal Access Token (PAT) or Service Account Token with `repo` and `issues: write` scope. |
+| `GOOGLE_API_KEY` | An API key for the Google AI (Gemini) model used for reasoning.                                      |
+
+### Optional Configuration
+
+These variables control the scanning behavior, thresholds, and model selection.
+
+| Variable Name          | Description                                                                                                        | Default                 |
+| :--------------------- | :----------------------------------------------------------------------------------------------------------------- | :---------------------- |
+| `INITIAL_FULL_SCAN`    | If `true`, audits every open issue in the repository. If `false`, only audits issues updated in the last 24 hours. | `false`                 |
+| `SPAM_LABEL_NAME`      | The exact text of the label applied to flagged issues.                                                             | `spam`                  |
+| `BOT_NAME`             | The GitHub username of your official bot to ensure its comments are ignored.                                       | `adk-bot`               |
+| `CONCURRENCY_LIMIT`    | The number of issues to process concurrently.                                                                      | `3`                     |
+| `SLEEP_BETWEEN_CHUNKS` | Time in seconds to sleep between batches to respect GitHub API rate limits.                                        | `1.5`                   |
+| `LLM_MODEL_NAME`       | The specific Gemini model version to use.                                                                          | `gemini-2.5-flash`      |
+| `OWNER`                | Repository owner (auto-detected in Actions).                                                                       | (Environment dependent) |
+| `REPO`                 | Repository name (auto-detected in Actions).                                                                        | (Environment dependent) |
+
+______________________________________________________________________
+
+## Deployment
+
+To deploy this agent, a GitHub Actions workflow file (`.github/workflows/issue-monitor.yml`) is recommended.
+
+### Directory Structure Note
+
+Because this agent resides within the `adk-python` package structure, the workflow must ensure the script is executed correctly to handle imports. It must be run as a module from the parent directory.
+
+### Example Workflow Execution
+
+```yaml
+      - name: Run ADK Issue Monitoring Agent
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          GOOGLE_API_KEY: ${{ secrets.GOOGLE_API_KEY }}
+          OWNER: ${{ github.repository_owner }}
+          REPO: ${{ github.event.repository.name }}
+          # Mapped to the manual trigger checkbox in the GitHub UI
+          INITIAL_FULL_SCAN: ${{ github.event.inputs.full_scan == 'true' }}
+          PYTHONPATH: contributing/samples
+        run: python -m adk_issue_monitoring_agent.main
+```

contributing/samples/adk_team/adk_issue_monitoring_agent/__init__.py

@@ -0,0 +1,15 @@
+# Copyright 2026 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from . import agent

contributing/samples/adk_team/adk_issue_monitoring_agent/agent.py

@@ -0,0 +1,118 @@
+# Copyright 2026 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import logging
+import os
+from typing import Any
+
+from adk_issue_monitoring_agent.settings import BOT_ALERT_SIGNATURE
+from adk_issue_monitoring_agent.settings import GITHUB_BASE_URL
+from adk_issue_monitoring_agent.settings import LLM_MODEL_NAME
+from adk_issue_monitoring_agent.settings import OWNER
+from adk_issue_monitoring_agent.settings import REPO
+from adk_issue_monitoring_agent.settings import SPAM_LABEL_NAME
+from adk_issue_monitoring_agent.utils import error_response
+from adk_issue_monitoring_agent.utils import get_issue_comments
+from adk_issue_monitoring_agent.utils import get_issue_details
+from adk_issue_monitoring_agent.utils import post_request
+from google.adk.agents.llm_agent import Agent
+from requests.exceptions import RequestException
+
+logger = logging.getLogger("google_adk." + __name__)
+
+
+def load_prompt_template(filename: str) -> str:
+  file_path = os.path.join(os.path.dirname(__file__), filename)
+  with open(file_path, "r") as f:
+    return f.read()
+
+
+PROMPT_TEMPLATE = load_prompt_template("PROMPT_INSTRUCTION.txt")
+
+# --- Tools ---
+
+
+def flag_issue_as_spam(
+    item_number: int, detection_reason: str
+) -> dict[str, Any]:
+  """
+  Flags an issue as spam by adding a label and leaving a comment for maintainers.
+  Includes idempotency checks to avoid duplicate POST actions.
+
+  Args:
+      item_number (int): The GitHub issue number.
+      detection_reason (str): The explanation of what the spam is.
+  """
+  logger.info(f"Flagging #{item_number} as SPAM. Reason: {detection_reason}")
+
+  label_url = (
+      f"{GITHUB_BASE_URL}/repos/{OWNER}/{REPO}/issues/{item_number}/labels"
+  )
+  comment_url = (
+      f"{GITHUB_BASE_URL}/repos/{OWNER}/{REPO}/issues/{item_number}/comments"
+  )
+
+  safe_reason = detection_reason.replace("```", "'''")
+
+  alert_body = (
+      f"{BOT_ALERT_SIGNATURE}\n"
+      "@maintainers, a suspected spam comment was detected in this thread.\n\n"
+      "**Reason:**\n"
+      f"```text\n{safe_reason}\n```"
+  )
+
+  try:
+    # 1. Fetch current state to check what actions are actually needed
+    issue = get_issue_details(OWNER, REPO, item_number)
+    comments = get_issue_comments(OWNER, REPO, item_number)
+
+    current_labels = [
+        label["name"].lower() for label in issue.get("labels", [])
+    ]
+    is_labeled = SPAM_LABEL_NAME.lower() in current_labels
+    is_commented = any(
+        BOT_ALERT_SIGNATURE in c.get("body", "") for c in comments
+    )
+
+    if is_labeled and is_commented:
+      logger.info(f"#{item_number} is already labeled and commented. Skipping.")
+    elif is_labeled and not is_commented:
+      post_request(comment_url, {"body": alert_body})
+      logger.info(f"Successfully posted spam alert comment to #{item_number}.")
+    elif not is_labeled and is_commented:
+      post_request(label_url, {"labels": [SPAM_LABEL_NAME]})
+      logger.info(
+          f"Successfully added '{SPAM_LABEL_NAME}' label to #{item_number}."
+      )
+    else:
+      post_request(label_url, {"labels": [SPAM_LABEL_NAME]})
+      post_request(comment_url, {"body": alert_body})
+      logger.info(f"Successfully fully flagged #{item_number}.")
+
+    return {"status": "success", "message": "Maintainers alerted successfully."}
+
+  except RequestException as e:
+    return error_response(f"Error flagging issue: {e}")
+
+
+root_agent = Agent(
+    model=LLM_MODEL_NAME,
+    name="spam_auditor_agent",
+    description="Audits issue comments for spam.",
+    instruction=PROMPT_TEMPLATE.format(
+        OWNER=OWNER,
+        REPO=REPO,
+    ),
+    tools=[flag_issue_as_spam],
+)

contributing/samples/adk_team/adk_issue_monitoring_agent/main.py

@@ -0,0 +1,204 @@
+# Copyright 2026 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import asyncio
+import logging
+import re
+import time
+
+from adk_issue_monitoring_agent.agent import root_agent
+from adk_issue_monitoring_agent.settings import BOT_ALERT_SIGNATURE
+from adk_issue_monitoring_agent.settings import BOT_NAME
+from adk_issue_monitoring_agent.settings import CONCURRENCY_LIMIT
+from adk_issue_monitoring_agent.settings import OWNER
+from adk_issue_monitoring_agent.settings import REPO
+from adk_issue_monitoring_agent.settings import SLEEP_BETWEEN_CHUNKS
+from adk_issue_monitoring_agent.utils import get_api_call_count
+from adk_issue_monitoring_agent.utils import get_issue_comments
+from adk_issue_monitoring_agent.utils import get_issue_details
+from adk_issue_monitoring_agent.utils import get_repository_maintainers
+from adk_issue_monitoring_agent.utils import get_target_issues
+from adk_issue_monitoring_agent.utils import reset_api_call_count
+from google.adk.cli.utils import logs
+from google.adk.runners import InMemoryRunner
+from google.genai import types
+
+logs.setup_adk_logger(level=logging.INFO)
+logger = logging.getLogger("google_adk." + __name__)
+
+APP_NAME = "issue_monitoring_app"
+USER_ID = "issue_monitoring_user"
+
+
+async def process_single_issue(
+    runner: InMemoryRunner, issue_number: int, maintainers: list[str]
+) -> tuple[float, int]:
+  start_time = time.perf_counter()
+  start_api_calls = get_api_call_count()
+
+  try:
+    # 1. Fetch the main issue AND the comments
+    issue = get_issue_details(OWNER, REPO, issue_number)
+    comments = get_issue_comments(OWNER, REPO, issue_number)
+
+    user_comments = []
+
+    # 2. Process the ORIGINAL ISSUE DESCRIPTION first!
+    issue_author = issue.get("user", {}).get("login", "")
+    issue_body = issue.get("body") or ""
+
+    # Only check the description if the author isn't a maintainer/bot
+    if (
+        issue_author not in maintainers
+        and not issue_author.endswith("[bot]")
+        and issue_author != BOT_NAME
+    ):
+      cleaned_issue_body = re.sub(
+          r"```.*?```", "\n[CODE BLOCK REMOVED]\n", issue_body, flags=re.DOTALL
+      )
+      if len(cleaned_issue_body) > 1500:
+        cleaned_issue_body = cleaned_issue_body[:1500] + "\n...[TRUNCATED]"
+      user_comments.append(
+          f"Author (Original Issue): @{issue_author}\nText:"
+          f" {cleaned_issue_body}\n---"
+      )
+
+    # 3. Process all the replies (comments)
+    for c in comments:
+      author = c.get("user", {}).get("login", "")
+      body = c.get("body") or ""
+
+      if BOT_ALERT_SIGNATURE in body:
+        logger.info(
+            f"#{issue_number}: Spam bot already alerted maintainers previously."
+            " Skipping."
+        )
+        return (
+            time.perf_counter() - start_time,
+            get_api_call_count() - start_api_calls,
+        )
+
+      if (
+          author in maintainers
+          or author.endswith("[bot]")
+          or author == BOT_NAME
+      ):
+        continue
+
+      cleaned_body = re.sub(
+          r"```.*?```", "\n[CODE BLOCK REMOVED]\n", body, flags=re.DOTALL
+      )
+
+      if len(cleaned_body) > 1500:
+        cleaned_body = cleaned_body[:1500] + "\n...[TRUNCATED]"
+
+      user_comments.append(f"Author: @{author}\nComment: {cleaned_body}\n---")
+
+    # 4. Skip LLM if no user text exists
+    if not user_comments:
+      logger.debug(f"#{issue_number}: No non-maintainer text found. Skipping.")
+      return (
+          time.perf_counter() - start_time,
+          get_api_call_count() - start_api_calls,
+      )
+
+    logger.info(
+        f"Processing Issue #{issue_number} (Found {len(user_comments)} items to"
+        " review)..."
+    )
+
+    # 5. Format prompt and invoke LLM
+    compiled_comments = "\n".join(user_comments)
+    prompt_text = (
+        "Please review the following text for issue"
+        f" #{issue_number}:\n\n{compiled_comments}"
+    )
+
+    session = await runner.session_service.create_session(
+        user_id=USER_ID, app_name=APP_NAME
+    )
+    prompt_message = types.Content(
+        role="user", parts=[types.Part(text=prompt_text)]
+    )
+
+    async for event in runner.run_async(
+        user_id=USER_ID, session_id=session.id, new_message=prompt_message
+    ):
+      if (
+          event.content
+          and event.content.parts
+          and hasattr(event.content.parts[0], "text")
+      ):
+        text = event.content.parts[0].text
+        if text:
+          clean_text = text[:100].replace("\n", " ")
+          logger.info(f"#{issue_number} Decision: {clean_text}...")
+
+  except Exception as e:
+    logger.error(f"Error processing issue #{issue_number}: {e}", exc_info=True)
+
+  # Calculate duration and API calls regardless of success or failure
+  duration = time.perf_counter() - start_time
+  issue_api_calls = get_api_call_count() - start_api_calls
+  return duration, issue_api_calls
+
+
+async def main():
+  logger.info(f"--- Starting Issue Monitoring Agent for {OWNER}/{REPO} ---")
+  reset_api_call_count()
+
+  # Step 1: Fetch Maintainers
+  try:
+    maintainers = get_repository_maintainers(OWNER, REPO)
+    logger.info(f"Found {len(maintainers)} maintainers.")
+  except Exception as e:
+    logger.critical(f"Failed to fetch maintainers: {e}")
+    return
+
+  # Step 2: Fetch target issues
+  try:
+    all_issues = get_target_issues(OWNER, REPO)
+  except Exception as e:
+    logger.critical(f"Failed to fetch issue list: {e}")
+    return
+
+  total_count = len(all_issues)
+  if total_count == 0:
+    logger.info("No issues matched criteria. Run finished.")
+    return
+
+  logger.info(f"Found {total_count} issues to process.")
+
+  # Initialize the runner ONCE for the entire run
+  runner = InMemoryRunner(agent=root_agent, app_name=APP_NAME)
+
+  # Step 3: Iterate through issues async 'CONCURRENCY_LIMIT' at a time
+  for i in range(0, total_count, CONCURRENCY_LIMIT):
+    chunk = all_issues[i : i + CONCURRENCY_LIMIT]
+    logger.info(f"Processing chunk: {chunk}")
+
+    tasks = [
+        process_single_issue(runner, issue_num, maintainers)
+        for issue_num in chunk
+    ]
+    await asyncio.gather(*tasks)
+
+    if (i + CONCURRENCY_LIMIT) < total_count:
+      await asyncio.sleep(SLEEP_BETWEEN_CHUNKS)
+
+  logger.info(f"--- Run Finished. Total API calls: {get_api_call_count()} ---")
+
+
+if __name__ == "__main__":
+  asyncio.run(main())

contributing/samples/adk_team/adk_issue_monitoring_agent/settings.py

@@ -0,0 +1,43 @@
+# Copyright 2026 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+from pathlib import Path
+
+from dotenv import load_dotenv
+
+CURRENT_DIR = Path(__file__).resolve().parent
+ENV_PATH = CURRENT_DIR / ".env"
+load_dotenv(dotenv_path=ENV_PATH, override=True)
+
+GITHUB_BASE_URL = "https://api.github.com"
+GITHUB_TOKEN = os.getenv("GITHUB_TOKEN")
+if not GITHUB_TOKEN:
+  raise ValueError("GITHUB_TOKEN environment variable not set")
+
+OWNER = os.getenv("OWNER", "google")
+REPO = os.getenv("REPO", "adk-python")
+LLM_MODEL_NAME = os.getenv("LLM_MODEL_NAME", "gemini-2.5-flash")
+
+SPAM_LABEL_NAME = os.getenv("SPAM_LABEL_NAME", "spam")
+CONCURRENCY_LIMIT = int(os.getenv("CONCURRENCY_LIMIT", 3))
+BOT_NAME = os.getenv("BOT_NAME", "adk-bot")
+BOT_ALERT_SIGNATURE = os.getenv(
+    "BOT_ALERT_SIGNATURE", "🚨 **Automated Spam Detection Alert** 🚨"
+)
+SLEEP_BETWEEN_CHUNKS = float(os.getenv("SLEEP_BETWEEN_CHUNKS", 1.5))
+
+
+# Toggle for the initial run
+INITIAL_FULL_SCAN = os.getenv("INITIAL_FULL_SCAN", "false").lower() == "true"

contributing/samples/adk_team/adk_issue_monitoring_agent/utils.py

@@ -0,0 +1,171 @@
+# Copyright 2026 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import logging
+from typing import Any
+
+from adk_issue_monitoring_agent.settings import GITHUB_TOKEN
+from adk_issue_monitoring_agent.settings import INITIAL_FULL_SCAN
+from adk_issue_monitoring_agent.settings import SPAM_LABEL_NAME
+import requests
+from requests.adapters import HTTPAdapter
+from urllib3.util.retry import Retry
+
+logger = logging.getLogger("google_adk." + __name__)
+
+_api_call_count = 0
+
+
+def get_api_call_count() -> int:
+  return _api_call_count
+
+
+def reset_api_call_count() -> None:
+  global _api_call_count
+  _api_call_count = 0
+
+
+def _increment_api_call_count() -> None:
+  global _api_call_count
+  _api_call_count += 1
+
+
+retry_strategy = Retry(
+    total=6,
+    backoff_factor=2,
+    status_forcelist=[429, 500, 502, 503, 504],
+    allowed_methods=["GET", "DELETE"],
+)
+adapter = HTTPAdapter(max_retries=retry_strategy)
+_session = requests.Session()
+_session.mount("https://", adapter)
+_session.headers.update({
+    "Authorization": f"token {GITHUB_TOKEN}",
+    "Accept": "application/vnd.github.v3+json",
+})
+
+
+def get_request(url: str, params: dict[str, Any] | None = None) -> Any:
+  _increment_api_call_count()
+  response = _session.get(url, params=params or {}, timeout=60)
+  response.raise_for_status()
+  return response.json()
+
+
+def post_request(url: str, payload: Any) -> Any:
+  _increment_api_call_count()
+  response = _session.post(url, json=payload, timeout=60)
+  response.raise_for_status()
+  return response.json()
+
+
+def error_response(error_message: str) -> dict[str, Any]:
+  return {"status": "error", "message": error_message}
+
+
+def get_repository_maintainers(owner: str, repo: str) -> list[str]:
+  """Fetches all users with push/maintain access."""
+  url = f"https://api.github.com/repos/{owner}/{repo}/collaborators"
+  data = get_request(url, {"permission": "push"})
+  return [user["login"] for user in data]
+
+
+def get_issue_details(
+    owner: str, repo: str, issue_number: int
+) -> dict[str, Any]:
+  """Fetches the main issue object to get the original description (body)."""
+  url = f"https://api.github.com/repos/{owner}/{repo}/issues/{issue_number}"
+  return get_request(url)
+
+
+def get_issue_comments(
+    owner: str, repo: str, issue_number: int
+) -> list[dict[str, Any]]:
+  """Fetches ALL comments for a specific issue, handling pagination."""
+  url = f"https://api.github.com/repos/{owner}/{repo}/issues/{issue_number}/comments"
+  all_comments = []
+  page = 1
+
+  while True:
+    data = get_request(url, params={"per_page": 100, "page": page})
+    if not data:
+      break
+
+    all_comments.extend(data)
+
+    if len(data) < 100:
+      break
+    page += 1
+
+  return all_comments
+
+
+def get_target_issues(owner: str, repo: str) -> list[int]:
+  """
+  Fetches issues.
+  If INITIAL_FULL_SCAN is True, fetches ALL open issues.
+  If False, fetches only issues updated in the last 24 hours using the 'since' parameter.
+  """
+  from datetime import datetime
+  from datetime import timedelta
+  from datetime import timezone
+
+  url = f"https://api.github.com/repos/{owner}/{repo}/issues"
+  params = {
+      "state": "open",
+      "per_page": 100,
+  }
+
+  if INITIAL_FULL_SCAN:
+    logger.info("INITIAL_FULL_SCAN is True. Fetching ALL open issues...")
+  else:
+    yesterday = (datetime.now(timezone.utc) - timedelta(days=1)).strftime(
+        "%Y-%m-%dT%H:%M:%SZ"
+    )
+    params["since"] = yesterday
+    logger.info(f"Daily mode: Fetching issues updated since {yesterday}...")
+
+  issue_numbers = []
+  page = 1
+
+  while True:
+    params["page"] = page
+    try:
+      items = get_request(url, params=params)
+
+      if not items:
+        break
+
+      for item in items:
+        if "pull_request" not in item:
+          # Extract all the label names on this issue
+          current_labels = [label["name"] for label in item.get("labels", [])]
+
+          # Only add the issue if it DOES NOT already have the spam label
+          if SPAM_LABEL_NAME not in current_labels:
+            issue_numbers.append(item["number"])
+          else:
+            logger.debug(
+                f"Skipping #{item['number']} - already marked as spam."
+            )
+
+      if len(items) < 100:
+        break
+
+      page += 1
+    except requests.exceptions.RequestException as e:
+      logger.error(f"Failed to fetch issues on page {page}: {e}")
+      break
+
+  return issue_numbers