fix(slack): NestBot deferred replies format without a second LLM call

[rajnisk]

Proposed change

Resolves #3693

This PR implements reply handling fixes for NestBot’s delayed auto-reply path (generate_ai_reply_if_unanswered): how answers are formatted and posted to Slack after the question-detector pipeline, without overlapping the Gemini provider, multi-agent orchestration, or interaction-UX PRs.

What changed

• Bug fix: generate_ai_reply_if_unanswered no longer calls get_blocks(ai_response_text, …), which incorrectly ran process_ai_query again on the model’s answer. It now uses format_blocks(), which only formats pre-rendered text.
• format_blocks() splits long answers into multiple Slack section blocks so mrkdwn stays under Slack’s per-block limit (~3001 characters; we target 3000).
• chat.postMessage text: fallback/notification text is derived from format_ai_response_for_slack, truncated with an ellipsis when needed, and aligned with block formatting (not raw model output only).
• format_ai_response_for_slack: stops calling format_links_for_slack directly; link conversion happens once via markdown() / format_links_for_slack per block (avoids double processing).
• Guards: reject bare YES / NO outputs from the pipeline in process_ai_query; treat empty/whitespace-only formatted output as error blocks instead of posting an empty block list.
• Tests: updated handler and auto-reply tests; added coverage for splitting, links, empty input, and YES/NO rejection.

Files touched

• backend/apps/slack/common/handlers/ai.py
• backend/apps/slack/services/message_auto_reply.py
• backend/apps/slack/utils.py
• backend/tests/apps/slack/common/handlers/ai_test.py
• backend/tests/apps/slack/services/message_auto_reply_test.py

Checklist

• [*] Required: I followed the contributing workflow
• [*] Required: I verified that my code works as intended and resolves the issue as described
• [*] Required: I ran make check-test locally: all warnings addressed, tests passed
• [*] I used AI for code, documentation, tests, or communication related to this PR

[coderabbitai]

Summary by CodeRabbit

Release Notes

• New Features

  • AI responses exceeding character limits are now split into multiple blocks for improved readability.

• Bug Fixes

  • Enhanced fallback text rendering in Slack messages with better formatting and truncation handling.
  • Improved edge case handling for empty or single-word AI responses.

Walkthrough

This PR refactors Slack AI response formatting by introducing a format_blocks() function that splits oversized AI responses into multiple valid Slack blocks with character limits. It updates process_ai_query() to filter out bare YES/NO responses, modifies format_ai_response_for_slack() to handle nullable input, and adjusts callers and tests throughout the Slack integration layer.

Changes

Cohort / File(s)	Summary
Core AI Handler backend/apps/slack/common/handlers/ai.py	Added MAX_BLOCK_TEXT_LENGTH constant and format_blocks() helper to split AI responses into multiple Slack blocks while respecting character limits. Updated get_blocks() to use format_blocks() instead of single markdown() block. Modified process_ai_query() to filter YES/NO responses, return None on exceptions, and add logging metadata.
Message Auto-Reply Service backend/apps/slack/services/message_auto_reply.py	Replaced get_blocks() call with format_blocks(). Added plain-text fallback with _FALLBACK_TEXT_MAX limit and whitespace handling, passing both blocks and text parameters to chat_postMessage().
Utilities backend/apps/slack/utils.py	Updated format_ai_response_for_slack() signature to accept str | None and return empty string for falsy inputs. Removed format_links_for_slack() call, delegating Markdown link conversion to per-block formatting elsewhere.
Test Coverage backend/tests/apps/slack/common/handlers/ai_test.py, backend/tests/apps/slack/services/message_auto_reply_test.py	Updated test mocks to patch format_blocks instead of get_blocks. Added new test cases for format_blocks() multi-block splitting, link rendering, and process_ai_query() YES/NO filtering. Updated test assertions for new function signatures and return types.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• #2113 — Modifies the same Slack AI handler and message auto-reply files with related block formatting and response handling changes
• #3939 — Updates process_ai_query() and imports in the same AI handler file with complementary logic modifications
• #2407 — Adds format_links_for_slack() utility that coordinates with this PR's refactoring of link conversion responsibility to per-block formatting

Suggested labels

nestbot, backend, backend-tests, Nestbot-ai-assistant

Suggested reviewers

• kasya
• arkid15r

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 inconclusive)

Check name	Status	Explanation	Resolution
Linked Issues check	❓ Inconclusive	The PR addresses reply formatting in the deferred auto-reply path but does not implement reaction management fixes, RAG improvements, or greeting routing corrections listed in issue #3693.	Verify whether this PR is intended as a partial fix scoped only to reply formatting, or if additional changes are needed to fully resolve issue #3693.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The PR title 'fix(slack): NestBot deferred replies format without a second LLM call' accurately describes the main change—eliminating a redundant LLM call in deferred reply handling and fixing reply formatting.
Description check	✅ Passed	The PR description comprehensively explains the changes, rationale, and files modified, clearly relating to the changeset and providing sufficient context.
Out of Scope Changes check	✅ Passed	All code changes are directly related to formatting replies and eliminating the duplicate LLM call in the deferred auto-reply path, with no unrelated modifications detected.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

📝 Coding Plan

• Generate coding plan for human review comments

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[cubic-dev-ai]

No issues found across 5 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

[coderabbitai]

🧹 Nitpick comments (1)

backend/tests/apps/slack/common/handlers/ai_test.py (1)

196-205: Consider using the constant for consistency.

The assertion len(block_text) <= 3001 hardcodes the Slack limit while MAX_BLOCK_TEXT_LENGTH (3000) is the production target. For clarity, consider either using <= MAX_BLOCK_TEXT_LENGTH to match the production guarantee, or adding a brief comment explaining that 3001 is Slack's actual limit while 3000 is the safety margin.

♻️ Suggested change

         for block in blocks:
             block_text = block["text"]["text"]
-            assert len(block_text) <= 3001  # Slack mrkdwn section limit
+            assert len(block_text) <= MAX_BLOCK_TEXT_LENGTH  # production safety limit

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@backend/tests/apps/slack/common/handlers/ai_test.py` around lines 196 - 205,
The test test_format_blocks_splits_long_text uses a hardcoded 3001 in the
assertion; update it to use the constant MAX_BLOCK_TEXT_LENGTH (or assert <=
MAX_BLOCK_TEXT_LENGTH + 1 with a clarifying comment if you intend to document
Slack's absolute limit) so the test aligns with the production target and
references the same limit used by format_blocks and the MAX_BLOCK_TEXT_LENGTH
constant.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@backend/tests/apps/slack/common/handlers/ai_test.py`:
- Around line 196-205: The test test_format_blocks_splits_long_text uses a
hardcoded 3001 in the assertion; update it to use the constant
MAX_BLOCK_TEXT_LENGTH (or assert <= MAX_BLOCK_TEXT_LENGTH + 1 with a clarifying
comment if you intend to document Slack's absolute limit) so the test aligns
with the production target and references the same limit used by format_blocks
and the MAX_BLOCK_TEXT_LENGTH constant.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: a83bbe0b-ceff-40e6-a584-bf03dfbd384e

📥 Commits

Reviewing files that changed from the base of the PR and between d6b2e85 and 712fb7f.

📒 Files selected for processing (5)

• backend/apps/slack/common/handlers/ai.py
• backend/apps/slack/services/message_auto_reply.py
• backend/apps/slack/utils.py
• backend/tests/apps/slack/common/handlers/ai_test.py
• backend/tests/apps/slack/services/message_auto_reply_test.py

[rajnisk]

Superseded: equivalent changes (deferred reply formatting without re-running the model, Slack block limits, fallback text, link handling, YES/NO + empty guards, tests) are merged into the #4319