[Feat] Discord 웹훅 로그 상세화 — 로드맵 생성/수정 파이프라인 내부 추적#70
Merged
Conversation
- job_start 로그로 생성/수정 요청 시작 시점 추적 - llm_router에서 모든 OpenAI API 호출 자동 추적 (stage, model, tier, latency) - 스켈레톤 생성 완료 시 슬롯 요약 및 repair/autofix 여부 포함 - Google Places 검색 시 fallback 단계별 통계 집계 - finalize 단계를 desc/time/summary 3단계로 분리 - 의도 분석 및 mutate 검색 상세 로그 추가 - 4000자 초과 시 detail 레벨 항목 자동 축약
- notify_callback_failure를 no-op으로 패치하여 call_count 오염 방지
13 tasks
Contributor
Author
|
@CodeRabbit review |
|
Note Reviews pausedUse the following commands to manage reviews:
Use the checkboxes below for quick actions:
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
✅ Files skipped from review due to trivial changes (1)
Walkthrough작업 로깅에 Changes
Sequence Diagram(s)mermaid Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related issues
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
✅ Actions performedReview triggered.
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
app/graph/roadmap/nodes/places.py (1)
147-165:⚠️ Potential issue | 🟡 Minor
region_bbox가 없는 슬롯은 지금fb_stage가 잘못 집계됩니다.
get_region_bbox(region)이None이면 첫 검색은 이미 unfiltered인데,fallback_stage기본값이"restriction"이라서 성공 시fb_stats에는 restriction hit로 누적되고places_searchdetail 로그도 건너뜁니다. 새 fallback 통계를 믿기 어렵게 되니 bbox 부재를 별도 stage로 분리하거나 초기값을 region 여부에 따라 설정하는 편이 맞습니다.예시 수정안
- fallback_stage = "restriction" + fallback_stage = "pending" restriction_used = False bias_used = False unfiltered_used = False region_bbox = get_region_bbox(region) if region_bbox is None: geo_missing_region_bbox = True + fallback_stage = "no_region_bbox" + else: + fallback_stage = "restriction"Also applies to: 265-271, 280-286, 355-359
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@app/graph/roadmap/nodes/places.py` around lines 147 - 165, In search_for_slot, when get_region_bbox(region) returns None the initial search is already unfiltered but fallback_stage is initialized to "restriction", causing fb_stats and places_search detail logs to misattribute the hit; change the initialization so fallback_stage reflects whether region_bbox is present (e.g., set fallback_stage = "unfiltered" or a distinct "missing_region" when region_bbox is None) and/or explicitly set fallback_stage immediately after computing region_bbox; update the same pattern in the other similar blocks (the later search branches that also call get_region_bbox and use fallback_stage/fb_stats) so that fb_stats counts and places_search detail logging use the correct stage when bbox is missing.
🧹 Nitpick comments (1)
app/core/job_log_context.py (1)
36-38:extra가 예약 필드를 덮어쓰지 못하게 막아 두는 편이 안전합니다.이번 PR부터
level이 로그 축약 동작을 제어하는 예약 필드가 됐는데, 지금은entry.update(extra)가 뒤에서 실행돼서extra={"level": ...}같은 값이 명시적level인자를 그대로 덮어쓸 수 있습니다.stage,message,elapsed_ms,level정도는 예약 키로 막아 두는 편이 안전합니다.예시 수정안
- entry: dict[str, Any] = {"stage": stage, "message": message, "elapsed_ms": elapsed_ms, "level": level} - if extra: - entry.update(extra) + entry: dict[str, Any] = {"stage": stage, "message": message, "elapsed_ms": elapsed_ms, "level": level} + if extra: + reserved = {"stage", "message", "elapsed_ms", "level"} + entry.update({key: value for key, value in extra.items() if key not in reserved})🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@app/core/job_log_context.py` around lines 36 - 38, The current code unconditionally does entry.update(extra) which allows callers to overwrite reserved fields; modify the code that builds and updates the log `entry` (the block that creates entry: dict[str, Any] = {"stage": stage, "message": message, "elapsed_ms": elapsed_ms, "level": level}) to filter `extra` so any keys in the reserved set {"stage", "message", "elapsed_ms", "level"} are dropped before merging (e.g. build a new dict from extra excluding those keys or pop those keys from extra) and then update `entry` with the filtered dict to ensure reserved fields cannot be overwritten.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@app/core/llm_router.py`:
- Around line 202-207: The llm_error paths currently log only the exception
type; update every error logging call that uses append_job_log with event
"llm_error" (and related error places noted around the append_job_log calls) to
include the same minimum common fields as the success path: include latency_ms
in the extra dict, include tier (use tier.value if tier else 'N/A') and fallback
(true/false) in the message or extra so webhook traces show which tier/fallback
path and latency caused the failure; keep the existing stage and selected_model
fields consistent with the success log format used by append_job_log("llm",
...).
In `@app/services/chat_service.py`:
- Around line 127-128: 현재 init_job_log와 append_job_log("job_start", f"type=chat
job_id={request.job_id} query={request.user_query[:60]}")에서 원문 user_query를 그대로
일부라도 남기고 있어 민감정보가 유출될 위험이 있습니다; 대신 request.user_query의 원문을 남기지 말고 길이(len), 안전한
해시(e.g., SHA256), 그리고 이미 파싱한 intent 또는 엔티티 요약만 로그에 기록하도록 변경하고(예:
append_job_log("job_start", f"type=chat job_id={request.job_id}
query_len={len(request.user_query)} query_hash={hash}
intent={intent_summary}")), 동일한 파일에서 사용된 다른 로그 패턴들(kw=, q= 계열 문자열 포함)을 찾아 같은
방식으로 교체하여 원문이 절대로 기록되지 않도록 하세요 (참조 심볼: init_job_log, append_job_log,
request.user_query, "job_start", any "kw=" / "q=" usages).
In `@app/services/google_places_service.py`:
- Around line 107-113: The detail log currently includes the raw query substring
(query[:50]) which may leak sensitive user text; update the append_job_log call
in google_places_service (around variables query, geo, price_tag, min_rating) to
omit the raw query and instead log safe metadata—e.g. include q_len (len(query))
or a short hash of query and the existing derived fields (geo, price_tag,
min_rating) so the original query text is never written to the detail log.
---
Outside diff comments:
In `@app/graph/roadmap/nodes/places.py`:
- Around line 147-165: In search_for_slot, when get_region_bbox(region) returns
None the initial search is already unfiltered but fallback_stage is initialized
to "restriction", causing fb_stats and places_search detail logs to misattribute
the hit; change the initialization so fallback_stage reflects whether
region_bbox is present (e.g., set fallback_stage = "unfiltered" or a distinct
"missing_region" when region_bbox is None) and/or explicitly set fallback_stage
immediately after computing region_bbox; update the same pattern in the other
similar blocks (the later search branches that also call get_region_bbox and use
fallback_stage/fb_stats) so that fb_stats counts and places_search detail
logging use the correct stage when bbox is missing.
---
Nitpick comments:
In `@app/core/job_log_context.py`:
- Around line 36-38: The current code unconditionally does entry.update(extra)
which allows callers to overwrite reserved fields; modify the code that builds
and updates the log `entry` (the block that creates entry: dict[str, Any] =
{"stage": stage, "message": message, "elapsed_ms": elapsed_ms, "level": level})
to filter `extra` so any keys in the reserved set {"stage", "message",
"elapsed_ms", "level"} are dropped before merging (e.g. build a new dict from
extra excluding those keys or pop those keys from extra) and then update `entry`
with the filtered dict to ensure reserved fields cannot be overwritten.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 3aeabd3a-c0cd-4b8b-963f-a0a321ce759b
📒 Files selected for processing (12)
app/core/job_log_context.pyapp/core/llm_router.pyapp/graph/chat/nodes/analyze_intent.pyapp/graph/chat/nodes/mutate.pyapp/graph/roadmap/nodes/finalize.pyapp/graph/roadmap/nodes/places.pyapp/graph/roadmap/nodes/skeleton.pyapp/services/chat_service.pyapp/services/generate_service.pyapp/services/google_places_service.pyapp/services/webhook_notification.pytests/test_callback_delivery.py
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@app/services/chat_service.py`:
- Around line 129-132: Replace the weak 8-char sha256 prefix used for query_hash
with an HMAC-based digest using a server-side secret and a longer prefix:
compute HMAC-SHA256 over request.user_query with a secret loaded from config/env
(fail loudly if missing), take a longer hex substring (e.g., 32 chars) and use
that value for query_hash in the append_job_log call; update references to
query_hash and ensure the secret access is performed securely (not checked into
source) so the change affects the calculation used in
append_job_log("job_start", ...).
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: fd3bdb93-7e34-40c6-b9dd-81f1bddc3f45
📒 Files selected for processing (3)
app/core/llm_router.pyapp/services/chat_service.pyapp/services/google_places_service.py
🚧 Files skipped from review as they are similar to previous changes (2)
- app/services/google_places_service.py
- app/core/llm_router.py
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@app/core/config.py`:
- Line 18: You made HMAC_SECRET a required setting (HMAC_SECRET in Settings), so
update the sample env file (.env.example) to include HMAC_SECRET with a clear
placeholder value and a short comment; add a line like
HMAC_SECRET=your_hmac_secret_here (or similar) so new deployments won’t fail
Settings initialization, and ensure any README or docs referencing required env
vars mention HMAC_SECRET as mandatory.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 26d1ace2-0039-4051-9c9c-cd6d3a90bd82
📒 Files selected for processing (8)
app/core/config.pyapp/services/chat_service.pyapp/services/google_places_service.pytests/test_callback_delivery.pytests/test_main.pytests/test_readiness.pytests/test_recommend_service.pytests/test_timeout_policy.py
✅ Files skipped from review due to trivial changes (2)
- tests/test_readiness.py
- tests/test_recommend_service.py
🚧 Files skipped from review as they are similar to previous changes (3)
- tests/test_callback_delivery.py
- app/services/chat_service.py
- app/services/google_places_service.py
Contributor
Author
|
@CodeRabbit ignore |
1 similar comment
Contributor
Author
|
@CodeRabbit ignore |
|
✅ Actions performedReviews paused. |
✅ Actions performedReviews paused. |
1 similar comment
✅ Actions performedReviews paused. |
iamb0ttle
approved these changes
Apr 8, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
1. 개요
2. 작업 내용
3. AI 활용 및 검증
AI가 생성한 코드 포함
100% 직접 작성
검증 방법: uv run pytest 35 passed
4. 스크린샷 (선택)
5. 체크리스트
uv run pre-commit run --all-files를 실행하여 통과했는가?Summary by CodeRabbit
새로운 기능
개선 사항
테스트
문서/환경