Implement Daily Civic Intelligence Refinement Engine#476
Implement Daily Civic Intelligence Refinement Engine#476RohanExploit wants to merge 3 commits intomainfrom
Conversation
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
✅ Deploy Preview for fixmybharat canceled.
|
🙏 Thank you for your contribution, @RohanExploit!PR Details:
Quality Checklist:
Review Process:
Note: The maintainers will monitor code quality and ensure the overall project flow isn't broken. |
📝 WalkthroughWalkthroughRedesigns CIVIC_INTELLIGENCE.md to a UTC-based daily workflow, introduces AdaptiveWeights-driven dynamic deduplication radius used in issue creation, fixes DBSCAN eps unit conversion (meters→radians) in spatial utilities, adds end-to-end tests for daily cycle, and updates runtime dependencies (numpy, scikit-learn). Changes
Sequence DiagramsequenceDiagram
participant Client as Issue Creator
participant Router as Issue Router
participant Weights as AdaptiveWeights
participant Spatial as Spatial Utils
participant DB as Database
participant Engine as CivicIntelligenceEngine
participant FS as File System
Client->>Router: Create issue payload
Router->>Weights: get_duplicate_search_radius()
Weights-->>Router: dynamic radius (meters)
Router->>Spatial: Compute bbox & query nearby within radius
Spatial->>DB: Query open issues in bbox
DB-->>Spatial: Nearby issues list
alt Nearby issues within radius
Spatial-->>Router: Duplicates detected (closest)
Router->>DB: Link report to existing issue
else No duplicates found
Router->>DB: Create new issue
end
Note over Engine: Daily cycle (UTC) runs
Engine->>DB: Fetch new issues, escalations
Engine->>Weights: Analyze trends, update weights
Weights->>FS: Persist modelWeights.json
Engine->>FS: Write daily snapshot (civic index, metrics)
FS-->>Engine: Snapshot & weights saved
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Suggested labels
Poem
🚥 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 |
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
Pull request overview
This PR implements a Daily Civic Intelligence Refinement Engine that automatically analyzes civic issues, detects trends, and optimizes system parameters on a daily schedule. The system learns from user behavior to improve duplicate detection accuracy and severity scoring.
Changes:
- Added system test for the full daily civic intelligence cycle
- Fixed DBSCAN clustering bug where radius was incorrectly converted from meters to degrees instead of radians
- Integrated dynamic duplicate detection radius from adaptive weights system into issue creation
Reviewed changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 4 comments.
| File | Description |
|---|---|
| backend/tests/test_civic_intelligence_system.py | New comprehensive system test for the daily refinement cycle |
| backend/spatial_utils.py | Fixed DBSCAN radius conversion from degrees to radians |
| backend/routers/issues.py | Integrated dynamic duplicate search radius from adaptive weights |
| CIVIC_INTELLIGENCE.md | Updated documentation with clearer structure and emoji formatting |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| # Patch paths to use temp directory | ||
| with patch("backend.adaptive_weights.DATA_FILE", weights_file), \ | ||
| patch("backend.civic_intelligence.SNAPSHOT_DIR", snapshots_dir), \ | ||
| patch("backend.civic_intelligence.SessionLocal", return_value=db_session): |
There was a problem hiding this comment.
The SessionLocal mock is incorrectly configured. SessionLocal is called as a factory function (SessionLocal()) in civic_intelligence.py line 45, so the mock needs to be callable and return the session. Import MagicMock from unittest.mock, then use: mock_session_local = MagicMock(return_value=db_session), patch("backend.civic_intelligence.SessionLocal", mock_session_local). Currently, the test passes db_session directly as return_value to patch, which will fail when SessionLocal() is called because db_session is a Session object, not a callable.
| new_pothole_weight = weights_system.get_category_multipliers()["pothole"] | ||
| assert new_pothole_weight > 1.0, f"Pothole weight should increase from 1.0, got {new_pothole_weight}" | ||
|
|
||
| # Radius should increase because of clustering (> 5 clusters) |
There was a problem hiding this comment.
The comment incorrectly states "Radius should increase because of clustering (> 5 clusters)" but the test setup creates scattered issues specifically to avoid forming clusters (line 73-74 comments). The radius increase is actually triggered by the "high volume, no clusters" condition (cluster_count == 0 and len(issues_24h) > 50) from civic_intelligence.py line 134-136. The comment should be corrected to reflect the actual test scenario.
| # Radius should increase because of clustering (> 5 clusters) | |
| # Radius should increase due to high volume of issues with no clusters (cluster_count == 0 and len(issues_24h) > 50) |
| from datetime import datetime, timedelta, timezone | ||
| from sqlalchemy import create_engine | ||
| from sqlalchemy.orm import sessionmaker | ||
| from unittest.mock import patch |
There was a problem hiding this comment.
Missing MagicMock import needed to fix the SessionLocal mocking pattern on line 60. Add MagicMock to the import statement: from unittest.mock import patch, MagicMock
| from unittest.mock import patch | |
| from unittest.mock import patch, MagicMock |
| # Convert eps from meters to radians | ||
| # Haversine metric expects inputs in radians and eps in radians | ||
| R = 6371000.0 # Earth's radius in meters | ||
| eps_radians = eps_meters / R |
There was a problem hiding this comment.
The DBSCAN radius conversion fix from degrees to radians lacks a direct unit test. While test_civic_intelligence_system.py implicitly tests clustering behavior, consider adding a unit test in test_spatial_utils.py that verifies cluster_issues_dbscan correctly clusters issues within the specified eps_meters distance. This would ensure the mathematical conversion is correct and prevent regression.
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
backend/routers/issues.py (1)
97-125:⚠️ Potential issue | 🟠 MajorValidate and clamp the adaptive dedup radius before using it.
At Line [98], the radius is consumed directly from adaptive weights. If that value is malformed (e.g., string/NaN/out-of-range), Line [102] or Line [125] can fail or misbehave, and the broad
exceptthen silently disables deduplication for that request.🔧 Suggested hardening
@@ import uuid import os import logging import hashlib +import math from datetime import datetime, timezone @@ - # Get dynamic search radius from adaptive weights - search_radius = adaptive_weights.get_duplicate_search_radius() + # Get dynamic search radius from adaptive weights + raw_radius = adaptive_weights.get_duplicate_search_radius() + try: + search_radius = float(raw_radius) + except (TypeError, ValueError): + logger.warning("Invalid duplicate_search_radius=%r; using fallback 50.0m", raw_radius) + search_radius = 50.0 + + if not math.isfinite(search_radius): + logger.warning("Non-finite duplicate_search_radius=%r; using fallback 50.0m", raw_radius) + search_radius = 50.0 + + # Keep runtime behavior bounded even if config file is edited manually + search_radius = max(10.0, min(search_radius, 200.0))🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@backend/routers/issues.py` around lines 97 - 125, Validate and clamp the value returned by adaptive_weights.get_duplicate_search_radius() before using it: ensure it's a finite number (coerce/parse if needed), fall back to a safe default when non-numeric, NaN, or missing, and clamp to a defined min/max (e.g., 10–5000 meters) to avoid extreme or invalid radii; then use that sanitized search_radius in get_bounding_box(...) and find_nearby_issues(...) and log a warning when you substitute a fallback so the run_in_threadpool/db.query dedup path doesn't silently disable deduplication.
🧹 Nitpick comments (1)
backend/tests/test_civic_intelligence_system.py (1)
98-98: Use_for intentionally unused loop variable.Line [98] uses
ibut never reads it; this is a straightforward lint cleanup.♻️ Small cleanup
- for i in range(4): # 4 upgrades > 3 threshold + for _ in range(4): # 4 upgrades > 3 threshold🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@backend/tests/test_civic_intelligence_system.py` at line 98, The loop in the test uses an unused loop variable "i" (for i in range(4):) — replace it with the conventional unused-name "_" (for _ in range(4):) to satisfy linting; update the occurrence in backend/tests/test_civic_intelligence_system.py where the loop is present (the for i in range(4): line) so the test behavior remains identical but the unused variable warning is resolved.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@backend/tests/test_civic_intelligence_system.py`:
- Around line 58-60: The test is leaking module-level adaptive_weights state
used by CivicIntelligenceEngine; instead of only patching
backend.adaptive_weights.DATA_FILE, replace or reset the module-level
adaptive_weights object so cached weights can't persist across tests—patch
backend.civic_intelligence.adaptive_weights to a fresh AdaptiveWeights instance
(or call its reset/clear method) before constructing CivicIntelligenceEngine,
ensuring DATA_FILE is still set to weights_file and SessionLocal/snapshot
patches remain in place.
In `@CIVIC_INTELLIGENCE.md`:
- Line 57: The docs list the snapshot field as `weight_changes` but the tests
assert `weight_updates`; update the snapshot schema so both match by renaming
the field in CIVIC_INTELLIGENCE.md from `weight_changes` to `weight_updates` (or
alternatively change the test to expect `weight_changes`), ensuring the field
name in the README and the schema/test (`weight_updates` vs `weight_changes`)
are identical across the codebase.
---
Outside diff comments:
In `@backend/routers/issues.py`:
- Around line 97-125: Validate and clamp the value returned by
adaptive_weights.get_duplicate_search_radius() before using it: ensure it's a
finite number (coerce/parse if needed), fall back to a safe default when
non-numeric, NaN, or missing, and clamp to a defined min/max (e.g., 10–5000
meters) to avoid extreme or invalid radii; then use that sanitized search_radius
in get_bounding_box(...) and find_nearby_issues(...) and log a warning when you
substitute a fallback so the run_in_threadpool/db.query dedup path doesn't
silently disable deduplication.
---
Nitpick comments:
In `@backend/tests/test_civic_intelligence_system.py`:
- Line 98: The loop in the test uses an unused loop variable "i" (for i in
range(4):) — replace it with the conventional unused-name "_" (for _ in
range(4):) to satisfy linting; update the occurrence in
backend/tests/test_civic_intelligence_system.py where the loop is present (the
for i in range(4): line) so the test behavior remains identical but the unused
variable warning is resolved.
ℹ️ Review info
Configuration used: defaults
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (4)
CIVIC_INTELLIGENCE.mdbackend/routers/issues.pybackend/spatial_utils.pybackend/tests/test_civic_intelligence_system.py
| with patch("backend.adaptive_weights.DATA_FILE", weights_file), \ | ||
| patch("backend.civic_intelligence.SNAPSHOT_DIR", snapshots_dir), \ | ||
| patch("backend.civic_intelligence.SessionLocal", return_value=db_session): |
There was a problem hiding this comment.
Isolate the adaptive-weights singleton to avoid cross-test state leakage.
Line [58]-Line [60] patches DATA_FILE, but CivicIntelligenceEngine uses a module-level adaptive_weights object with cached state. That can make this test order-dependent if cached weights survive into later tests.
🧪 Suggested isolation patch
@@
def test_daily_civic_intelligence_cycle(temp_dirs, db_session):
- temp_dir, weights_file, snapshots_dir = temp_dirs
+ _, weights_file, snapshots_dir = temp_dirs
+ test_weights = AdaptiveWeights()
@@
- with patch("backend.adaptive_weights.DATA_FILE", weights_file), \
- patch("backend.civic_intelligence.SNAPSHOT_DIR", snapshots_dir), \
- patch("backend.civic_intelligence.SessionLocal", return_value=db_session):
+ with patch("backend.adaptive_weights.DATA_FILE", weights_file), \
+ patch("backend.civic_intelligence.SNAPSHOT_DIR", snapshots_dir), \
+ patch("backend.civic_intelligence.SessionLocal", return_value=db_session), \
+ patch("backend.civic_intelligence.adaptive_weights", test_weights):
@@
- weights_system = AdaptiveWeights()
- weights_system._weights = None # Force reload
- weights_system._load_weights()
+ test_weights._weights = None # Force reload from patched DATA_FILE
+ test_weights._last_loaded = 0
+ test_weights._load_weights()
@@
- assert weights_system.get_category_multipliers()["pothole"] == 1.0
- assert weights_system.get_duplicate_search_radius() == 50.0
+ assert test_weights.get_category_multipliers()["pothole"] == 1.0
+ assert test_weights.get_duplicate_search_radius() == 50.0
@@
- weights_system._last_loaded = 0 # Force reload
- weights_system._load_weights()
+ test_weights._last_loaded = 0 # Force reload
+ test_weights._load_weights()
@@
- new_pothole_weight = weights_system.get_category_multipliers()["pothole"]
+ new_pothole_weight = test_weights.get_category_multipliers()["pothole"]
@@
- new_radius = weights_system.get_duplicate_search_radius()
+ new_radius = test_weights.get_duplicate_search_radius()📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| with patch("backend.adaptive_weights.DATA_FILE", weights_file), \ | |
| patch("backend.civic_intelligence.SNAPSHOT_DIR", snapshots_dir), \ | |
| patch("backend.civic_intelligence.SessionLocal", return_value=db_session): | |
| def test_daily_civic_intelligence_cycle(temp_dirs, db_session): | |
| _, weights_file, snapshots_dir = temp_dirs | |
| test_weights = AdaptiveWeights() | |
| with patch("backend.adaptive_weights.DATA_FILE", weights_file), \ | |
| patch("backend.civic_intelligence.SNAPSHOT_DIR", snapshots_dir), \ | |
| patch("backend.civic_intelligence.SessionLocal", return_value=db_session), \ | |
| patch("backend.civic_intelligence.adaptive_weights", test_weights): | |
| test_weights._weights = None # Force reload from patched DATA_FILE | |
| test_weights._last_loaded = 0 | |
| test_weights._load_weights() | |
| assert test_weights.get_category_multipliers()["pothole"] == 1.0 | |
| assert test_weights.get_duplicate_search_radius() == 50.0 | |
| # Simulate weight update | |
| test_weights._last_loaded = 0 # Force reload | |
| test_weights._load_weights() | |
| new_pothole_weight = test_weights.get_category_multipliers()["pothole"] | |
| new_radius = test_weights.get_duplicate_search_radius() |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@backend/tests/test_civic_intelligence_system.py` around lines 58 - 60, The
test is leaking module-level adaptive_weights state used by
CivicIntelligenceEngine; instead of only patching
backend.adaptive_weights.DATA_FILE, replace or reset the module-level
adaptive_weights object so cached weights can't persist across tests—patch
backend.civic_intelligence.adaptive_weights to a fresh AdaptiveWeights instance
(or call its reset/clear method) before constructing CivicIntelligenceEngine,
ensuring DATA_FILE is still set to weights_file and SessionLocal/snapshot
patches remain in place.
| They contain: | ||
| * `trends`: Top keywords, category distribution, clusters. | ||
| * `civic_index`: The daily score and insights. | ||
| * `weight_changes`: Audit log of any automatic weight adjustments. |
There was a problem hiding this comment.
Snapshot field name is inconsistent with tested payload.
Line [57] documents weight_changes, but backend/tests/test_civic_intelligence_system.py (Line [138]-Line [139]) validates weight_updates. Please align the doc with the actual snapshot schema (or vice versa).
📝 Doc fix
-* `weight_changes`: Audit log of any automatic weight adjustments.
+* `weight_updates`: Audit log of any automatic weight adjustments.📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| * `weight_changes`: Audit log of any automatic weight adjustments. | |
| * `weight_updates`: Audit log of any automatic weight adjustments. |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@CIVIC_INTELLIGENCE.md` at line 57, The docs list the snapshot field as
`weight_changes` but the tests assert `weight_updates`; update the snapshot
schema so both match by renaming the field in CIVIC_INTELLIGENCE.md from
`weight_changes` to `weight_updates` (or alternatively change the test to expect
`weight_changes`), ensuring the field name in the README and the schema/test
(`weight_updates` vs `weight_changes`) are identical across the codebase.
🔍 Quality Reminder |
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 `@backend/requirements-render.txt`:
- Around line 23-24: The new requirements entries add unpinned packages; pin
scikit-learn (and numpy if you intend to control it) to specific versions to
ensure reproducible Render deployments: update the requirements-render.txt
entries for "scikit-learn" (and "numpy" if kept) to explicit version pins (e.g.,
scikit-learn==X.Y.Z and numpy==A.B.C) or remove the explicit numpy line if you
want numpy to follow scikit-learn's transitive constraint; modify the lines
referencing scikit-learn and numpy accordingly so they are fixed to chosen
versions.
backend/requirements-render.txt
Outdated
| scikit-learn | ||
| numpy |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
# First, check if the file exists and read the relevant section
cat -n backend/requirements-render.txtRepository: RohanExploit/VishwaGuru
Length of output: 557
🌐 Web query:
scikit-learn latest version 2025 PyPI
💡 Result:
On PyPI, the latest scikit-learn release in 2025 (and the latest stable version listed) is 1.8.0, published Dec 10, 2025. [1]
🌐 Web query:
numpy latest version 2025 PyPI
💡 Result:
On PyPI, the latest NumPy release during 2025 (i.e., the newest one with a 2025 release date) was NumPy 2.3.5, released November 16, 2025. [1]
(For context: PyPI shows later releases in 2026, e.g. 2.4.2 on January 31, 2026.) [1]
Sources: [1] PyPI “numpy” project page (release history).
🌐 Web query:
scikit-learn numpy transitive dependency
💡 Result:
-
In scikit-learn itself, NumPy is a direct (runtime) dependency, i.e., scikit-learn declares that it requires NumPy (along with SciPy, joblib, threadpoolctl). The scikit-learn docs list NumPy as a required “build, install” dependency (e.g., minimum versions are published in their dependency table). [1]
-
For your project, NumPy is often a transitive dependency if you only depend on scikit-learn. Example: if your project declares
scikit-learnbut notnumpy, thennumpyis an indirect requirement that arrives because scikit-learn needs it—so it’s transitive from your project’s perspective. pip explicitly resolves “dependencies of dependencies” during installation. [2] -
Also note:
numpydoes not depend onscikit-learn(the dependency direction is scikit-learn → numpy, not the other way around). [1]
Practical checks / control:
- See the dependency tree in an environment with
pipdeptree(e.g.,pip install pipdeptree && pipdeptree -p scikit-learn). [3] - If you need to pin or restrict a transitive dependency (like NumPy pulled in via scikit-learn/SciPy), pip recommends using a constraints file. [2]
Sources:
[1] scikit-learn docs (install + minimum dependency table)
[2] pip docs (dependency resolution; constraints for transitive deps)
[3] pipdeptree (PyPI description/usage)
🌐 Web query:
scikit-learn CVE security advisories 2025
💡 Result:
scikit-learn security advisories / CVEs in 2025
- Upstream scikit-learn (the
scikit-learnproject itself): GitHub’s Security Advisories page forscikit-learn/scikit-learnshows no published security advisories, i.e., no GHSA entries to map to 2025 CVEs/advisories. [1] - Known scikit-learn CVEs around that time (but not 2025): Vulnerability trackers for the PyPI package list items such as CVE-2024-5206 (fixed in scikit-learn 1.5.0) and older issues like CVE-2020-28975, but do not show any CVE-2025-… entries for scikit-learn itself. [2]
What you may see in 2025 anyway (downstream/vendor advisories)
- Linux distros may publish 2025-dated advisories for their
python*-scikit-learnpackages that reference older CVEs (example: an openSUSE 2025 advisory referencing CVE-2020-28975). [3]
Closely related 2025 advisory (scikit-learn ecosystem)
- CVE-2025-54886 affects
skops(used to share/ship scikit-learn models), enabling arbitrary code execution in certain loading paths; fixed in skops 0.13.0. This is not a scikit-learn CVE, but is relevant if you use skops with scikit-learn models. [4]
Sources: [1] GitHub scikit-learn security advisories page, [2] Safety vulnerability listing for scikit-learn, [3] openSUSE 2025 advisory, [4] CVE-2025-54886 details (Wiz).
🌐 Web query:
numpy CVE security advisories 2025
💡 Result:
NumPy CVE / security advisories in 2025
-
NumPy upstream (project-maintained advisories): As of Feb 25, 2026, the official
numpy/numpyGitHub repository shows no published security advisories (so there were none in 2025). [1] -
CVE entries in 2025 that some distro errata associate with “numpy” updates are typically not NumPy bugs, but ecosystem / bundle updates where the NumPy RPM/wheel gets rebuilt alongside other Python components:
- CVE-2025-8194 is a CPython
tarfileinfinite-loop/deadlock issue (not NumPy). [2] Oracle Linux errata (Aug 29, 2025) lists a “python39” security update where numpy is among updated packages while referencing CVE-2025-8194. [3] - CVE-2025-47273 is a setuptools path traversal issue (not NumPy). [4] Some third-party vulnerability trackers/distro rollups list numpy among affected/updated packages for this CVE as part of broader module updates. [4]
- CVE-2025-8194 is a CPython
If you tell me your environment (pip vs conda vs OS packages, and which NumPy version), I can point to the specific advisories/errata that apply to your installation.
Sources: [1] [2] [3] [4]
Pin versions for both new dependencies to ensure reproducible Render deployments.
Both scikit-learn and numpy are added unpinned. Because Render resolves the latest satisfying version on every deploy, a major or breaking release of either package will silently break the next deployment without any code change. This is the same risk as all the other unpinned entries in this file, but it's worth addressing here since these are newly introduced.
Additionally, numpy is a hard transitive dependency of scikit-learn; explicitly listing it without a pin adds noise but no constraint — only worth keeping if you need to enforce a specific numpy version.
📌 Suggested version pinning
-scikit-learn
-numpy
+scikit-learn==1.8.0
+numpy==2.4.2📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| scikit-learn | |
| numpy | |
| scikit-learn==1.8.0 | |
| numpy==2.4.2 |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@backend/requirements-render.txt` around lines 23 - 24, The new requirements
entries add unpinned packages; pin scikit-learn (and numpy if you intend to
control it) to specific versions to ensure reproducible Render deployments:
update the requirements-render.txt entries for "scikit-learn" (and "numpy" if
kept) to explicit version pins (e.g., scikit-learn==X.Y.Z and numpy==A.B.C) or
remove the explicit numpy line if you want numpy to follow scikit-learn's
transitive constraint; modify the lines referencing scikit-learn and numpy
accordingly so they are fixed to chosen versions.
Implements the Daily Civic Intelligence Refinement Engine that runs daily to analyze civic issues, detect trends, and optimize system parameters (severity weights and duplicate detection radius) automatically.
backend/routers/issues.pyto use dynamic duplicate search radius fromAdaptiveWeights.backend/spatial_utils.pywhere DBSCAN radius was calculated in degrees but passed to a metric expecting radians.backend/tests/test_civic_intelligence_system.pycovering the full daily cycle.CIVIC_INTELLIGENCE.md.PR created automatically by Jules for task 18289678482842708093 started by @RohanExploit
Summary by cubic
Implements a daily engine that analyzes civic issues, learns from admin upgrades, and auto-tunes severity weights and duplicate detection radius. Integrates the adaptive radius into issue creation, fixes DBSCAN units for haversine, and optimizes Render dependencies.
New Features
Dependencies
Written for commit af4ed8f. Summary will update on new commits.
Summary by CodeRabbit
New Features
Bug Fixes
Documentation
Tests
Chores