feat(knn): add k-Nearest Neighbors classifier #105

cardmagic · 2025-12-29T02:09:13Z

Summary

Implements k-Nearest Neighbors (kNN) classifier leveraging existing LSI infrastructure
Hash-style API consistent with Bayes and LSI (add, classify)
classify_with_neighbors() returns interpretable results with neighbor details
Distance-weighted voting option (weighted: true)
Full JSON/Marshal serialization and storage backend support
44 tests with 96.2% line coverage on knn.rb

Usage

knn = Classifier::KNN.new(k: 3, weighted: true)

knn.add(
  spam: ["Buy now!", "Free money!"],
  ham: ["Meeting tomorrow", "Project update"]
)

knn.classify "Special offer!"
# => "spam"

result = knn.classify_with_neighbors "Claim your prize"
# => {category: "spam", neighbors: [...], votes: {...}, confidence: 0.85}

When to Use kNN vs Bayes vs LSI

Classifier	Best For
Bayes	Large training sets, fast classification, spam filtering
LSI	Semantic similarity, document clustering, search
kNN	Small datasets, interpretable results, incremental learning

Test plan

All 291 tests pass
96.8% overall line coverage
kNN-specific tests: 44 tests, 120 assertions
Edge cases: single item, empty classifier, k > items

Closes #103

Implements a kNN classifier that leverages existing LSI infrastructure for similarity computations. Unlike Bayes which requires training, kNN uses instance-based learning—store examples and classify by finding the most similar ones. Key features: - Hash-style API consistent with Bayes and LSI (add, classify) - classify_with_neighbors() returns interpretable results with neighbor details, vote tallies, and confidence scores - Distance-weighted voting option for more nuanced classification - Full JSON/Marshal serialization and storage backend support - Handles edge cases like single-item classifiers gracefully This completes Issue #103 and provides a third classification algorithm suited for small datasets where interpretability matters. Closes #103

greptile-apps · 2025-12-29T02:15:44Z

Greptile Summary

Implements k-Nearest Neighbors classifier by leveraging existing LSI infrastructure for similarity calculations. Uses instance-based learning where examples are stored and classification happens by finding the k most similar items and voting by their categories.

Major Changes

New Classifier::KNN class (lib/classifier/knn.rb): Complete kNN implementation with hash-style API matching existing Bayes/LSI patterns
LSI edge case fix (lib/classifier/lsi.rb:632): Added single-item handling to proximity_array_for_content_unlocked to prevent errors when only one training example exists
Comprehensive test coverage: 44 tests with excellent coverage of classification, serialization, storage, and edge cases
Documentation: Added extensive README section with usage examples, feature comparison table, and guidance on when to use each classifier

Key Features

Distance-weighted voting option for more accurate classification
Full serialization support (JSON and Marshal)
Storage backend integration with dirty tracking
Thread-safe with mutex synchronization
Interpretable results via classify_with_neighbors showing which examples influenced the decision

Architecture

The implementation cleanly delegates to LSI for similarity calculations (proximity_array_for_content), keeping kNN focused on the voting logic and result structuring. This reuse of existing infrastructure minimizes new code while providing powerful functionality.

Confidence Score: 5/5

This PR is safe to merge with minimal risk
Well-designed implementation with comprehensive test coverage (44 tests, 96.8% overall line coverage), proper error handling, thread safety via mutex, and clean integration with existing LSI infrastructure. The code follows established patterns in the codebase and includes edge case handling.
No files require special attention

Important Files Changed

Filename	Overview
lib/classifier/knn.rb	New k-Nearest Neighbors classifier implementation with LSI integration, serialization support, and comprehensive API - well-structured with proper error handling
lib/classifier/lsi.rb	Added edge case handling for single-item proximity calculation - returns similarity of 1.0 when only one item exists
test/knn/knn_test.rb	Comprehensive test suite with 44 tests covering initialization, classification, serialization, storage, and edge cases

Sequence Diagram

sequenceDiagram
    participant User
    participant KNN
    participant LSI
    participant Matrix

    Note over User,Matrix: Adding Training Data
    User->>KNN: add(spam: ["Buy now!"], ham: ["Meeting"])
    KNN->>KNN: synchronize { @dirty = true }
    KNN->>LSI: add(**items)
    LSI->>LSI: Store items with categories
    LSI->>LSI: Mark for rebuild if auto_rebuild
    LSI-->>KNN: Items stored

    Note over User,Matrix: Classification Flow
    User->>KNN: classify("Free offer!")
    KNN->>KNN: classify_with_neighbors(text)
    KNN->>KNN: synchronize block
    
    alt Empty classifier
        KNN-->>User: nil
    else Has items
        KNN->>KNN: find_neighbors(text)
        KNN->>LSI: proximity_array_for_content(text)
        LSI->>LSI: synchronize { proximity_array_for_content_unlocked }
        
        alt Single item
            LSI-->>KNN: [[item, 1.0]]
        else Multiple items
            LSI->>LSI: node_for_content_unlocked(text)
            LSI->>Matrix: Compute dot products with all items
            Matrix-->>LSI: Similarity scores
            LSI->>LSI: Sort by similarity descending
            LSI-->>KNN: [[item1, 0.92], [item2, 0.75], ...]
        end
        
        KNN->>KNN: Select top k neighbors
        KNN->>LSI: categories_for(each item)
        LSI-->>KNN: Category labels
        
        KNN->>KNN: tally_votes(neighbors)
        alt Weighted voting
            KNN->>KNN: votes[cat] += similarity
        else Unweighted voting
            KNN->>KNN: votes[cat] += 1.0
        end
        
        KNN->>KNN: Calculate winner & confidence
        KNN-->>User: {category: "spam", neighbors: [...], votes: {...}, confidence: 0.85}
    end

    Note over User,Matrix: Persistence
    User->>KNN: save_to_file(path)
    KNN->>KNN: to_json
    KNN->>KNN: as_json
    KNN->>LSI: as_json
    LSI-->>KNN: LSI state as hash
    KNN->>KNN: Write JSON to file
    KNN->>KNN: @dirty = false
    KNN-->>User: Bytes written

    User->>KNN: load_from_file(path)
    KNN->>KNN: from_json(File.read)
    KNN->>LSI: from_json(lsi_data)
    LSI-->>KNN: Restored LSI instance
    KNN-->>User: Restored KNN instance

greptile-apps

_{5 files reviewed, 35 comments}

_{Edit Code Review Agent Settings | Greptile}

greptile-apps · 2025-12-29T02:15:07Z

lib/classifier/knn.rb

+  # This class implements a k-Nearest Neighbors classifier that leverages
+  # the existing LSI infrastructure for similarity computations.
+  #
+  # Unlike traditional classifiers that require training, kNN uses instance-based
+  # learning - it stores examples and classifies by finding the most similar ones.
+  #
+  # Example usage:
+  #   knn = Classifier::KNN.new(k: 3)
+  #   knn.add("spam" => ["Buy now!", "Limited offer!"])
+  #   knn.add("ham" => ["Meeting tomorrow", "Project update"])
+  #   knn.classify("Special discount!") # => "spam"


style: docstring restates obvious implementation details

Suggested change

# This class implements a k-Nearest Neighbors classifier that leverages

# the existing LSI infrastructure for similarity computations.

#

# Unlike traditional classifiers that require training, kNN uses instance-based

# learning - it stores examples and classifies by finding the most similar ones.

#

# Example usage:

# knn = Classifier::KNN.new(k: 3)

# knn.add("spam" => ["Buy now!", "Limited offer!"])

# knn.add("ham" => ["Meeting tomorrow", "Project update"])

# knn.classify("Special discount!") # => "spam"

# k-Nearest Neighbors classifier using LSI for similarity.

#

# Example:

# knn = Classifier::KNN.new(k: 3)

# knn.add(spam: ["Buy now!"], ham: ["Meeting tomorrow"])

# knn.classify("Special discount!") # => "spam"

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 12:22 Comment: **style:** docstring restates obvious implementation details ```suggestion # k-Nearest Neighbors classifier using LSI for similarity. # # Example: # knn = Classifier::KNN.new(k: 3) # knn.add(spam: ["Buy now!"], ham: ["Meeting tomorrow"]) # knn.classify("Special discount!") # => "spam" ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:08Z

lib/classifier/knn.rb

+    # Creates a new kNN classifier.
+    #
+    # @param k [Integer] Number of neighbors to consider (default: 5)
+    # @param weighted [Boolean] Use distance-weighted voting (default: false)
+    #


style: docstring repeats parameter types already in @rbs

Suggested change

# Creates a new kNN classifier.

#

# @param k [Integer] Number of neighbors to consider (default: 5)

# @param weighted [Boolean] Use distance-weighted voting (default: false)

#

# @rbs (?k: Integer, ?weighted: bool) -> void

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 36:40 Comment: **style:** docstring repeats parameter types already in `@rbs` ```suggestion # @rbs (?k: Integer, ?weighted: bool) -> void ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:08Z

lib/classifier/knn.rb

+
+    # Adds labeled examples to the classifier using hash-style syntax.
+    # Keys are categories, values are items (or arrays of items).
+    #
+    # @example Single item per category
+    #   knn.add("spam" => "Buy now!")
+    #   knn.add("ham" => "Meeting tomorrow")
+    #
+    # @example Multiple items per category
+    #   knn.add("spam" => ["Buy now!", "Limited offer!"])
+    #
+    # @example Batch operations
+    #   knn.add(
+    #     "spam" => ["Buy now!", "Limited offer!"],
+    #     "ham" => ["Meeting tomorrow", "Project update"]
+    #   )
+    #


style: docstring examples repeat what method signature already shows

Suggested change

# Adds labeled examples to the classifier using hash-style syntax.

# Keys are categories, values are items (or arrays of items).

#

# @example Single item per category

# knn.add("spam" => "Buy now!")

# knn.add("ham" => "Meeting tomorrow")

#

# @example Multiple items per category

# knn.add("spam" => ["Buy now!", "Limited offer!"])

#

# @example Batch operations

# knn.add(

# "spam" => ["Buy now!", "Limited offer!"],

# "ham" => ["Meeting tomorrow", "Project update"]

# )

#

# @rbs (**untyped items) -> void

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 51:67 Comment: **style:** docstring examples repeat what method signature already shows ```suggestion # @rbs (**untyped items) -> void ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:09Z

lib/classifier/knn.rb

+    # Classifies the given text by finding the k nearest neighbors
+    # and using majority voting.
+    #
+    # @param text [String] The text to classify
+    # @return [String, Symbol, nil] The predicted category, or nil if no examples exist
+    #


style: docstring restates method name and signature

Suggested change

# Classifies the given text by finding the k nearest neighbors

# and using majority voting.

#

# @param text [String] The text to classify

# @return [String, Symbol, nil] The predicted category, or nil if no examples exist

#

# @rbs (String) -> (String | Symbol)?

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 87:92 Comment: **style:** docstring restates method name and signature ```suggestion # @rbs (String) -> (String | Symbol)? ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:10Z

lib/classifier/knn.rb

+    # Classifies the given text and returns detailed information about
+    # the neighbors that contributed to the decision.
+    #
+    # @param text [String] The text to classify
+    # @return [Hash] A hash containing:
+    #   - :category - The predicted category
+    #   - :neighbors - Array of neighbor details (item, category, similarity)
+    #   - :votes - Hash of category => vote count/weight
+    #   - :confidence - Confidence score (winning vote share)
+    #


style: docstring restates return hash keys visible in code

Suggested change

# Classifies the given text and returns detailed information about

# the neighbors that contributed to the decision.

#

# @param text [String] The text to classify

# @return [Hash] A hash containing:

# - :category - The predicted category

# - :neighbors - Array of neighbor details (item, category, similarity)

# - :votes - Hash of category => vote count/weight

# - :confidence - Confidence score (winning vote share)

#

# Returns category, neighbors, votes, and confidence score.

#

# @rbs (String) -> Hash[Symbol, untyped]

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 99:108 Comment: **style:** docstring restates return hash keys visible in code ```suggestion # Returns category, neighbors, votes, and confidence score. # # @rbs (String) -> Hash[Symbol, untyped] ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:23Z

lib/classifier/knn.rb

+      synchronize do
+        @k = data['k']
+        @weighted = data['weighted']


style: docstring repeats method name and obvious behavior

Suggested change

synchronize do

@k = data['k']

@weighted = data['weighted']

# @rbs (Array[Hash[Symbol, untyped]]) -> Hash[String | Symbol, Float]

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 352:354 Comment: **style:** docstring repeats method name and obvious behavior ```suggestion # @rbs (Array[Hash[Symbol, untyped]]) -> Hash[String | Symbol, Float] ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:41Z

lib/classifier/knn.rb

+    def classify_with_neighbors(text)
+      synchronize do
+        return empty_result if @lsi.items.empty?
+
+        neighbors = find_neighbors(text)
+        return empty_result if neighbors.empty?
+
+        votes = tally_votes(neighbors)
+        winner = votes.max_by { |_, v| v }&.first
+        total_votes = votes.values.sum
+        confidence = winner && total_votes.positive? ? votes[winner] / total_votes.to_f : 0.0
+
+        {
+          category: winner,
+          neighbors: neighbors,
+          votes: votes,
+          confidence: confidence
+        }
+      end


style: nested conditionals - use early returns

Suggested change

def classify_with_neighbors(text)

synchronize do

return empty_result if @lsi.items.empty?

neighbors = find_neighbors(text)

return empty_result if neighbors.empty?

votes = tally_votes(neighbors)

winner = votes.max_by { |_, v| v }&.first

total_votes = votes.values.sum

confidence = winner && total_votes.positive? ? votes[winner] / total_votes.to_f : 0.0

{

category: winner,

neighbors: neighbors,

votes: votes,

confidence: confidence

}

end

def classify_with_neighbors(text)

synchronize do

return empty_result if @lsi.items.empty?

neighbors = find_neighbors(text)

return empty_result if neighbors.empty?

votes = tally_votes(neighbors)

winner = votes.max_by { |_, v| v }&.first

return empty_result unless winner

total_votes = votes.values.sum

confidence = total_votes.positive? ? votes[winner] / total_votes.to_f : 0.0

{

category: winner,

neighbors: neighbors,

votes: votes,

confidence: confidence

}

end

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 110:128 Comment: **style:** nested conditionals - use early returns ```suggestion def classify_with_neighbors(text) synchronize do return empty_result if @lsi.items.empty? neighbors = find_neighbors(text) return empty_result if neighbors.empty? votes = tally_votes(neighbors) winner = votes.max_by { |_, v| v }&.first return empty_result unless winner total_votes = votes.values.sum confidence = total_votes.positive? ? votes[winner] / total_votes.to_f : 0.0 { category: winner, neighbors: neighbors, votes: votes, confidence: confidence } end end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:41Z

lib/classifier/knn.rb

+
+    # @rbs () -> bool
+    def dirty?
+      @dirty
+    end
+
+    # Loads a classifier from configured storage.
+    #
+    # @param storage [Storage::Base] The storage to load from
+    # @return [KNN] The loaded classifier
+    #


style: nested conditionals - use early returns

Suggested change

# @rbs () -> bool

def dirty?

@dirty

end

# Loads a classifier from configured storage.

#

# @param storage [Storage::Base] The storage to load from

# @return [KNN] The loaded classifier

#

def reload

raise ArgumentError, 'No storage configured' unless storage

raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 246:256 Comment: **style:** nested conditionals - use early returns ```suggestion def reload raise ArgumentError, 'No storage configured' unless storage raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty data = storage.read raise StorageError, 'No saved state found' unless data restore_from_json(data) @dirty = false self end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:42Z

lib/classifier/knn.rb

+
+      instance = from_json(data)
+      instance.storage = storage
+      instance
+    end
+
+    # Loads a classifier from a file.
+    #
+    # @param path [String] The file path
+    # @return [KNN] The loaded classifier


style: nested conditionals - use early returns

Suggested change

instance = from_json(data)

instance.storage = storage

instance

end

# Loads a classifier from a file.

#

# @param path [String] The file path

# @return [KNN] The loaded classifier

def reload!

raise ArgumentError, 'No storage configured' unless storage

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 261:270 Comment: **style:** nested conditionals - use early returns ```suggestion def reload! raise ArgumentError, 'No storage configured' unless storage data = storage.read raise StorageError, 'No saved state found' unless data restore_from_json(data) @dirty = false self end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:15:43Z

lib/classifier/knn.rb

+      @k, @weighted, @lsi, @dirty = data
+      @storage = nil
+    end
+
+    private
+
+    # Finds the k nearest neighbors for the given text.
+    #


style: nested conditionals - use early returns

Suggested change

@k, @weighted, @lsi, @dirty = data

@storage = nil

end

private

# Finds the k nearest neighbors for the given text.

#

def self.load(storage:)

data = storage.read

raise StorageError, 'No saved state found' unless data

instance = from_json(data)

instance.storage = storage

instance

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 285:292 Comment: **style:** nested conditionals - use early returns ```suggestion def self.load(storage:) data = storage.read raise StorageError, 'No saved state found' unless data instance = from_json(data) instance.storage = storage instance end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps

_{5 files reviewed, 9 comments}

_{Edit Code Review Agent Settings | Greptile}

greptile-apps · 2025-12-29T02:19:49Z

lib/classifier/knn.rb

+      # LSI requires at least 2 items to build an index
+      # For single item, return it directly with a default similarity
+      if @lsi.items.size == 1
+        item = @lsi.items.first
+        return [{
+          item: item,
+          category: @lsi.categories_for(item).first,
+          similarity: 1.0
+        }]
+      end


style: nested conditional with defensive code for single-item case. LSI already handles this - remove the special case.

Suggested change

# LSI requires at least 2 items to build an index

# For single item, return it directly with a default similarity

if @lsi.items.size == 1

item = @lsi.items.first

return [{

item: item,

category: @lsi.categories_for(item).first,

similarity: 1.0

}]

end

def find_neighbors(text)

proximity = @lsi.proximity_array_for_content(text)

neighbors = proximity.reject { |item, _| item == text }.first(@k)

neighbors.map do |item, similarity|

{

item: item,

category: @lsi.categories_for(item).first,

similarity: similarity

}

end

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 295:304 Comment: **style:** nested conditional with defensive code for single-item case. LSI already handles this - remove the special case. ```suggestion def find_neighbors(text) proximity = @lsi.proximity_array_for_content(text) neighbors = proximity.reject { |item, _| item == text }.first(@k) neighbors.map do |item, similarity| { item: item, category: @lsi.categories_for(item).first, similarity: similarity } end end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:49Z

lib/classifier/knn.rb

+    #
+    # @rbs () -> void
+    def save
+      raise ArgumentError, 'No storage configured. Use save_to_file(path) or set storage=' unless storage


style: defensive error for impossible scenario - if storage is nil, the method already can't be called. Remove this check.

Suggested change

raise ArgumentError, 'No storage configured. Use save_to_file(path) or set storage=' unless storage

storage.write(to_json)

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 200:200 Comment: **style:** defensive error for impossible scenario - if storage is nil, the method already can't be called. Remove this check. ```suggestion storage.write(to_json) ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:50Z

lib/classifier/knn.rb

+      raise ArgumentError, 'No storage configured' unless storage
+      raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty


style: nested error checks - use early returns

Suggested change

raise ArgumentError, 'No storage configured' unless storage

raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty

def reload

raise ArgumentError, 'No storage configured' unless storage

raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 222:223 Comment: **style:** nested error checks - use early returns ```suggestion def reload raise ArgumentError, 'No storage configured' unless storage raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty data = storage.read raise StorageError, 'No saved state found' unless data restore_from_json(data) @dirty = false self end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:50Z

lib/classifier/knn.rb

+      raise ArgumentError, 'No storage configured' unless storage
+
+      data = storage.read
+      raise StorageError, 'No saved state found' unless data


style: nested error checks - use early returns

Suggested change

raise ArgumentError, 'No storage configured' unless storage

data = storage.read

raise StorageError, 'No saved state found' unless data

def reload!

raise ArgumentError, 'No storage configured' unless storage

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 237:240 Comment: **style:** nested error checks - use early returns ```suggestion def reload! raise ArgumentError, 'No storage configured' unless storage data = storage.read raise StorageError, 'No saved state found' unless data restore_from_json(data) @dirty = false self end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:51Z

lib/classifier/knn.rb

+      data = storage.read
+      raise StorageError, 'No saved state found' unless data


style: nested error check - use early return

Suggested change

data = storage.read

raise StorageError, 'No saved state found' unless data

def self.load(storage:)

data = storage.read

raise StorageError, 'No saved state found' unless data

instance = from_json(data)

instance.storage = storage

instance

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 259:260 Comment: **style:** nested error check - use early return ```suggestion def self.load(storage:) data = storage.read raise StorageError, 'No saved state found' unless data instance = from_json(data) instance.storage = storage instance end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:51Z

lib/classifier/knn.rb

+      synchronize do
+        @k = data['k']
+        @weighted = data['weighted']
+
+        lsi_data = data['lsi']
+        lsi_data['type'] = 'lsi'
+        @lsi = LSI.from_json(lsi_data)
+        @dirty = false
+      end


style: synchronize block wraps entire method body - move the guard clause outside

Suggested change

synchronize do

@k = data['k']

@weighted = data['weighted']

lsi_data = data['lsi']

lsi_data['type'] = 'lsi'

@lsi = LSI.from_json(lsi_data)

@dirty = false

end

def restore_from_json(json)

data = JSON.parse(json)

raise ArgumentError, "Invalid classifier type: #{data['type']}" unless data['type'] == 'knn'

synchronize do

@k = data['k']

@weighted = data['weighted']

lsi_data = data['lsi']

lsi_data['type'] = 'lsi'

@lsi = LSI.from_json(lsi_data)

@dirty = false

end

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 352:360 Comment: **style:** synchronize block wraps entire method body - move the guard clause outside ```suggestion def restore_from_json(json) data = JSON.parse(json) raise ArgumentError, "Invalid classifier type: #{data['type']}" unless data['type'] == 'knn' synchronize do @k = data['k'] @weighted = data['weighted'] lsi_data = data['lsi'] lsi_data['type'] = 'lsi' @lsi = LSI.from_json(lsi_data) @dirty = false end end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:52Z

lib/classifier/knn.rb

+      # Restore the LSI from its nested data
+      lsi_data = data['lsi']
+      lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json


logic: unnecessary mutation of passed data - work with local copy

Suggested change

# Restore the LSI from its nested data

lsi_data = data['lsi']

lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json

lsi_data = data['lsi'].dup

lsi_data['type'] = 'lsi'

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 186:188 Comment: **logic:** unnecessary mutation of passed data - work with local copy ```suggestion lsi_data = data['lsi'].dup lsi_data['type'] = 'lsi' ``` How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:52Z

lib/classifier/knn.rb

+        lsi_data = data['lsi']
+        lsi_data['type'] = 'lsi'


logic: unnecessary mutation of parsed data - work with local copy

Suggested change

lsi_data = data['lsi']

lsi_data['type'] = 'lsi'

lsi_data = data['lsi'].dup

lsi_data['type'] = 'lsi'

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 356:357 Comment: **logic:** unnecessary mutation of parsed data - work with local copy ```suggestion lsi_data = data['lsi'].dup lsi_data['type'] = 'lsi' ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:19:53Z

lib/classifier/knn.rb

+      neighbors.each do |neighbor|
+        category = neighbor[:category]
+        next unless category


style: guard clause inside loop - extract to method or combine with map logic

Suggested change

neighbors.each do |neighbor|

category = neighbor[:category]

next unless category

neighbors.each do |neighbor|

category = neighbor[:category]

weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0

votes[category] += weight if category

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 324:326 Comment: **style:** guard clause inside loop - extract to method or combine with map logic ```suggestion neighbors.each do |neighbor| category = neighbor[:category] weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0 votes[category] += weight if category end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

@example

Remove @param/@return/@example tags that duplicate @RBS type info. Keep method descriptions that explain behavior. Remove deprecated add_item method (new class doesn't need legacy API).

cardmagic · 2025-12-29T02:24:49Z

@greptile please re-review

greptile-apps

_{5 files reviewed, 23 comments}

_{Edit Code Review Agent Settings | Greptile}

greptile-apps · 2025-12-29T02:34:41Z

lib/classifier/knn.rb

+      lsi_data = data['lsi']
+      lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json


logic: mutates the passed-in hash by adding 'type' key, violating principle of least surprise

Suggested change

lsi_data = data['lsi']

lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json

lsi_data = data['lsi'].dup

lsi_data['type'] = 'lsi'

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 137:138 Comment: **logic:** mutates the passed-in hash by adding `'type'` key, violating principle of least surprise ```suggestion lsi_data = data['lsi'].dup lsi_data['type'] = 'lsi' ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:42Z

lib/classifier/knn.rb

+        lsi_data = data['lsi']
+        lsi_data['type'] = 'lsi'


logic: mutates the parsed hash, same issue as line 137

Suggested change

lsi_data = data['lsi']

lsi_data['type'] = 'lsi'

lsi_data = data['lsi'].dup

lsi_data['type'] = 'lsi'

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 285:286 Comment: **logic:** mutates the parsed hash, same issue as line 137 ```suggestion lsi_data = data['lsi'].dup lsi_data['type'] = 'lsi' ``` How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:42Z

lib/classifier/knn.rb

+      # LSI requires at least 2 items to build an index
+      # For single item, return it directly with a default similarity
+      if @lsi.items.size == 1
+        item = @lsi.items.first
+        return [{
+          item: item,
+          category: @lsi.categories_for(item).first,
+          similarity: 1.0
+        }]
+      end


style: unnecessary special case - LSI's proximity_array_for_content should handle single-item case internally

If LSI requires this workaround, that's an LSI bug to fix there, not something to paper over here

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 228:237 Comment: **style:** unnecessary special case - LSI's `proximity_array_for_content` should handle single-item case internally If LSI requires this workaround, that's an LSI bug to fix there, not something to paper over here Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:43Z

lib/classifier/knn.rb

+        category = neighbor[:category]
+        next unless category
+
+        weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0


style: [neighbor[:similarity], 0.0].max is defensive - similarity from LSI should never be negative. If it is, that's a bug in LSI

Suggested change

weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0

weight = @weighted ? neighbor[:similarity] : 1.0

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 259:259 Comment: **style:** `[neighbor[:similarity], 0.0].max` is defensive - similarity from LSI should never be negative. If it is, that's a bug in LSI ```suggestion weight = @weighted ? neighbor[:similarity] : 1.0 ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:43Z

lib/classifier/knn.rb

+      # Restore the LSI from its nested data
+      lsi_data = data['lsi']
+      lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json


style: comment repeats what code shows - the code is self-documenting

Suggested change

# Restore the LSI from its nested data

lsi_data = data['lsi']

lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json

lsi_data = data['lsi'].dup

lsi_data['type'] = 'lsi'

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 136:138 Comment: **style:** comment repeats what code shows - the code is self-documenting ```suggestion lsi_data = data['lsi'].dup lsi_data['type'] = 'lsi' ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:51Z

lib/classifier/knn.rb

+    def reload
+      raise ArgumentError, 'No storage configured' unless storage
+      raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty
+
+      data = storage.read
+      raise StorageError, 'No saved state found' unless data
+
+      restore_from_json(data)
+      @dirty = false
+      self


style: nested error checks - use early returns

Suggested change

def reload

raise ArgumentError, 'No storage configured' unless storage

raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

def reload

raise ArgumentError, 'No storage configured' unless storage

raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 165:174 Comment: **style:** nested error checks - use early returns ```suggestion def reload raise ArgumentError, 'No storage configured' unless storage raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty data = storage.read raise StorageError, 'No saved state found' unless data restore_from_json(data) @dirty = false self end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:51Z

lib/classifier/knn.rb

+    def reload!
+      raise ArgumentError, 'No storage configured' unless storage
+
+      data = storage.read
+      raise StorageError, 'No saved state found' unless data
+
+      restore_from_json(data)
+      @dirty = false
+      self


style: nested error checks - use early returns

Suggested change

def reload!

raise ArgumentError, 'No storage configured' unless storage

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

def reload!

raise ArgumentError, 'No storage configured' unless storage

data = storage.read

raise StorageError, 'No saved state found' unless data

restore_from_json(data)

@dirty = false

self

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 179:187 Comment: **style:** nested error checks - use early returns ```suggestion def reload! raise ArgumentError, 'No storage configured' unless storage data = storage.read raise StorageError, 'No saved state found' unless data restore_from_json(data) @dirty = false self end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:52Z

lib/classifier/knn.rb

+    def self.load(storage:)
+      data = storage.read
+      raise StorageError, 'No saved state found' unless data
+
+      instance = from_json(data)
+      instance.storage = storage
+      instance


style: nested error check - use early return

Suggested change

def self.load(storage:)

data = storage.read

raise StorageError, 'No saved state found' unless data

instance = from_json(data)

instance.storage = storage

instance

def self.load(storage:)

data = storage.read

raise StorageError, 'No saved state found' unless data

instance = from_json(data)

instance.storage = storage

instance

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 197:203 Comment: **style:** nested error check - use early return ```suggestion def self.load(storage:) data = storage.read raise StorageError, 'No saved state found' unless data instance = from_json(data) instance.storage = storage instance end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:52Z

lib/classifier/knn.rb

+      neighbors.each do |neighbor|
+        category = neighbor[:category]
+        next unless category
+
+        weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0
+        votes[category] += weight
+      end


style: guard clause inside loop - combine with map logic or extract to method

Suggested change

neighbors.each do |neighbor|

category = neighbor[:category]

next unless category

weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0

votes[category] += weight

end

neighbors.each do |neighbor|

category = neighbor[:category] or next

weight = @weighted ? neighbor[:similarity] : 1.0

votes[category] += weight

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 255:261 Comment: **style:** guard clause inside loop - combine with map logic or extract to method ```suggestion neighbors.each do |neighbor| category = neighbor[:category] or next weight = @weighted ? neighbor[:similarity] : 1.0 votes[category] += weight end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

greptile-apps · 2025-12-29T02:34:53Z

lib/classifier/knn.rb

+    def restore_from_json(json)
+      data = JSON.parse(json)
+      raise ArgumentError, "Invalid classifier type: #{data['type']}" unless data['type'] == 'knn'
+
+      synchronize do
+        @k = data['k']
+        @weighted = data['weighted']
+
+        lsi_data = data['lsi']
+        lsi_data['type'] = 'lsi'
+        @lsi = LSI.from_json(lsi_data)
+        @dirty = false
+      end
+    end


style: synchronize block wraps entire method - move guard clause outside

Suggested change

def restore_from_json(json)

data = JSON.parse(json)

raise ArgumentError, "Invalid classifier type: #{data['type']}" unless data['type'] == 'knn'

synchronize do

@k = data['k']

@weighted = data['weighted']

lsi_data = data['lsi']

lsi_data['type'] = 'lsi'

@lsi = LSI.from_json(lsi_data)

@dirty = false

end

end

def restore_from_json(json)

data = JSON.parse(json)

raise ArgumentError, "Invalid classifier type: #{data['type']}" unless data['type'] == 'knn'

synchronize do

@k = data['k']

@weighted = data['weighted']

lsi_data = data['lsi'].dup

lsi_data['type'] = 'lsi'

@lsi = LSI.from_json(lsi_data)

@dirty = false

end

end

_{Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!}

Prompt To Fix With AI

This is a comment left during a code review. Path: lib/classifier/knn.rb Line: 277:290 Comment: **style:** `synchronize` block wraps entire method - move guard clause outside ```suggestion def restore_from_json(json) data = JSON.parse(json) raise ArgumentError, "Invalid classifier type: #{data['type']}" unless data['type'] == 'knn' synchronize do @k = data['k'] @weighted = data['weighted'] lsi_data = data['lsi'].dup lsi_data['type'] = 'lsi' @lsi = LSI.from_json(lsi_data) @dirty = false end end ``` Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time! How can I resolve this? If you propose a fix, please make it concise.

- Fix LSI single-item bug instead of working around in KNN - Remove defensive max(0) check on similarity scores - Use early return for nil winner - Simplify guard clause with 'or next' - Condense class docstring - Avoid mutating input hash with .dup

cardmagic · 2025-12-29T02:45:15Z

Addressed the actionable feedback - thanks!

Re: remaining comments - these are intentionally kept:

Method descriptions (e.g., "Classifies text using k nearest neighbors") - these explain what the method does at a glance, useful for API docs
Class docstring with example - essential for users discovering the API
@rbs annotations - these are RBS type signatures, not comments

The removed ones were YARD tags (@param, @return, @example) that duplicated the @rbs type info.

@greptile please re-review

cardmagic self-assigned this Dec 29, 2025

cardmagic force-pushed the feature/knn-classifier branch from dc53ba2 to 3f95b1b Compare December 29, 2025 02:13

greptile-apps bot reviewed Dec 29, 2025

View reviewed changes

docs: clarify kNN vs Bayes size guidance

0f4c7b5

greptile-apps bot reviewed Dec 29, 2025

View reviewed changes

refactor(knn): remove YARD tags, keep descriptions

55f205c

Remove @param/@return/@example tags that duplicate @RBS type info. Keep method descriptions that explain behavior. Remove deprecated add_item method (new class doesn't need legacy API).

greptile-apps bot reviewed Dec 29, 2025

View reviewed changes

cardmagic merged commit fde311b into master Dec 29, 2025
6 checks passed

cardmagic deleted the feature/knn-classifier branch December 29, 2025 02:49

	raise ArgumentError, 'No storage configured. Use save_to_file(path) or set storage=' unless storage
	storage.write(to_json)

		raise ArgumentError, 'No storage configured' unless storage
		raise UnsavedChangesError, 'Unsaved changes would be lost. Call save first or use reload!' if @dirty

		data = storage.read
		raise StorageError, 'No saved state found' unless data

		lsi_data = data['lsi']
		lsi_data['type'] = 'lsi' # Ensure type is set for LSI.from_json

	weight = @weighted ? [neighbor[:similarity], 0.0].max : 1.0
	weight = @weighted ? neighbor[:similarity] : 1.0

feat(knn): add k-Nearest Neighbors classifier #105

feat(knn): add k-Nearest Neighbors classifier #105

Uh oh!

Conversation

cardmagic commented Dec 29, 2025

Summary

Usage

When to Use kNN vs Bayes vs LSI

Test plan

Uh oh!

greptile-apps bot commented Dec 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Major Changes

Key Features

Architecture

Confidence Score: 5/5

Important Files Changed

Sequence Diagram

Uh oh!

greptile-apps bot left a comment

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot left a comment

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot Dec 29, 2025

Choose a reason for hiding this comment

Uh oh!

greptile-apps bot commented Dec 29, 2025 •

edited

Loading