Skip to content

chore: Streamline CodeRabbit config to only non-default values #142

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jeremyeder merged 1 commit into from
Nov 25, 2025
Merged

chore: Streamline CodeRabbit config to only non-default values #142

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
249 changes: 81 additions & 168 deletions .coderabbit.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,99 +1,31 @@
# CodeRabbit Configuration for AgentReady
# Goal: Extremely high signal-to-noise ratio for AI code reviews
# Last Updated: 2025-11-25
# Only non-default values specified for maximum clarity
# Full schema: https://coderabbit.ai/integrations/schema.v2.json

# Language and tone settings
language: en-US
early_access: false
enable_free_tier: true

# Review behavior - optimized for signal
# Disable poems (default: true)
reviews:
# Use "chill" profile to reduce noise
profile: "chill"

# Request changes only for critical issues
request_changes_workflow: true

# High-level overview without verbose walkthrough
high_level_summary: true
high_level_summary_placeholder: "<!-- CODERABBIT SUMMARY -->"

# Disable poem (pure noise)
poem: false

# Collapse walkthrough to reduce visual clutter
collapse_walkthrough: true

# Review status at top of PR for quick scanning
review_status: true

# Auto-review configuration
auto_review:
enabled: true
drafts: false # Don't waste cycles on WIP
base_branches:
- main
# Request changes for critical issues (default: false)
request_changes_workflow: true

# Tools configuration - only high-signal checks
# Disable tools we run locally to reduce noise
tools:
# Linting tools (we already run these locally, skip duplicates)
# Python formatting/linting (we run locally before push)
ruff:
enabled: false # We run ruff locally before push
black:
enabled: false # We run black locally before push
isort:
enabled: false # We run isort locally before push

# GitHub Actions analysis (useful for CI/CD changes)
actionlint:
enabled: true

# Shellcheck for bash scripts
shellcheck:
enabled: true
enabled: false # default: true
flake8:
enabled: false # default: true
pylint:
enabled: false # default: true

# Markdownlint (we run locally but good backup)
markdownlint:
enabled: true

# Gitleaks for secret detection (high signal)
gitleaks:
enabled: true

# Biome (JavaScript/TypeScript - not applicable but harmless)
# We don't use these technologies
biome:
enabled: false

# Hadolint for Dockerfiles (future use)
hadolint:
enabled: true

# What to focus on (HIGH SIGNAL ONLY)
focus:
- security_vulnerabilities # Critical: SQL injection, XSS, etc.
- bug_risks # High: Logic errors, race conditions
- error_handling # High: Uncaught exceptions, poor error messages
- breaking_changes # High: API compatibility
- performance_issues # Medium: Only significant issues (O(n²) → O(n))
- test_coverage_gaps # Medium: Missing critical test cases
- documentation_gaps # Low: Only for public APIs and complex logic

# What to ignore (NOISE REDUCTION)
ignore:
- style_preferences # Handled by black/isort/ruff locally
- naming_conventions # Handled by ruff locally
- line_length # We explicitly don't enforce this
- minor_refactoring_suggestions # Only suggest if major impact
- subjective_improvements # "This could be cleaner" noise
- trivial_optimizations # Micro-optimizations with no real impact

# Severity thresholds - only comment on important issues
severity_threshold: "medium" # Skip "low" severity comments

# Path-based rules
enabled: false # default: true (JS/TS tool, not applicable)

# Path-specific review instructions
path_instructions:
# Core library code - highest scrutiny
# Core assessors - enforce patterns
- path: "src/agentready/assessors/**"
instructions: |
- Verify BaseAssessor pattern compliance
Expand All @@ -102,132 +34,113 @@ reviews:
- Validate attribute_id matches research report
- Check for graceful degradation when tools missing

# Data models - check for breaking changes
- path: "src/agentready/models/**"
instructions: |
- Check for breaking changes to data models
- Verify backwards compatibility
- Ensure schema version bumps when needed

# Services - security and performance focus
- path: "src/agentready/services/**"
instructions: |
- Check for proper error handling
- Verify performance implications (file I/O, subprocess calls)
- Look for security issues (path traversal, command injection)

# Tests - focus on coverage and correctness
# Tests - coverage and correctness only
- path: "tests/**"
instructions: |
- Verify test actually tests the intended behavior
- Check for missing edge cases
- Flag overly brittle tests (mocking too much)
- Skip style comments entirely
- Skip all style comments

# CLI - focus on UX and error messages
# CLI - UX and error messages
- path: "src/agentready/cli/**"
instructions: |
- Check for clear error messages
- Verify help text is accurate
- Look for missing error handling for user input

# GitHub workflows - focus on security and correctness
# GitHub Actions - security focus
- path: ".github/workflows/**"
instructions: |
- Check for secret exposure risks
- Verify proper permissions (least privilege)
- Flag outdated action versions (security)

# Documentation - only flag critical issues
# Documentation - critical issues only
- path: "**/*.md"
instructions: |
- Only comment on factual errors or broken links
- Skip formatting/style suggestions
- Flag outdated information (version mismatches, wrong commands)

# Scripts - focus on security and robustness
# Scripts - security and robustness
- path: "scripts/**"
instructions: |
- Check for command injection vulnerabilities
- Verify error handling for subprocess calls
- Flag missing input validation

# Path filters - completely skip reviewing these
path_filters:
# Build artifacts and caches
- "!**/*.pyc"
- "!**/__pycache__/**"
- "!.venv/**"
- "!venv/**"
- "!htmlcov/**"
- "!.pytest_cache/**"
- "!.ruff_cache/**"
- "!*.egg-info/**"
- "!build/**"
- "!dist/**"

# Generated reports and data
- "!.agentready/**"
- "!.cache/**"
- "!*.log"
- "!*.tmp"

# Lock files (dependency updates are separate concern)
- "!uv.lock"
- "!poetry.lock"
- "!package-lock.json"

# Example outputs (reference only, not production code)
- "!examples/**/*.json"
- "!examples/**/*.html"
- "!examples/**/*.md"

# Custom instructions - AgentReady-specific context
chat:
auto_reply: true

# Knowledge base - teach CodeRabbit about AgentReady patterns
# Skip reviewing these paths entirely
path_filters:
- "!**/*.pyc"
- "!**/__pycache__/**"
- "!.venv/**"
- "!venv/**"
- "!htmlcov/**"
- "!.pytest_cache/**"
- "!.ruff_cache/**"
- "!*.egg-info/**"
- "!build/**"
- "!dist/**"
- "!.agentready/**"
- "!.cache/**"
- "!*.log"
- "!*.tmp"
- "!uv.lock"
- "!poetry.lock"
- "!package-lock.json"
- "!examples/**/*.json"
- "!examples/**/*.html"
- "!examples/**/*.md"

# Teach CodeRabbit about AgentReady-specific patterns
knowledge_base:
- |
AgentReady is a tool that assesses repositories against 25 agent-ready best practices.
The research report (agent-ready-codebase-attributes.md) defines these attributes.
All assessors must inherit from BaseAssessor and implement attribute_id property and assess() method.

- |
We intentionally don't enforce line length limits (E501 ignored in ruff).
We use black for formatting with default settings.
We prefer explicit over clever, simple over complex.

- |
Error handling philosophy: Assessors should return Finding.create_skipped() when tools are missing,
not crash. Fail gracefully and provide actionable remediation guidance.

- |
This project follows conventional commits (feat:, fix:, chore:, docs:, test:, refactor:).
All commits should be squashed before merge.
All commits must be signed with git signature.

- |
Testing philosophy: We aim for >80% coverage on new code.
Tests should focus on behavior, not implementation details.
Mock external dependencies, but don't over-mock internal functions.

- |
Documentation: CLAUDE.md is the source of truth for development.
README.md is user-facing. Keep them in sync.
All new features must update CLAUDE.md before merging.

# Review thresholds - only create review comments for actionable items
review:
# Minimum confidence to comment (0.0 - 1.0)
# Higher = fewer but more accurate comments
confidence_threshold: 0.75

# Require at least this many similar patterns before suggesting refactoring
pattern_threshold: 3

# Only suggest performance improvements if >10% impact
performance_improvement_threshold: 0.10

# Tone and style - professional and concise
code_guidelines:
# Custom guidelines (default has empty array)
learnings:
- |
AgentReady assesses repositories against 25 agent-ready best practices.
Research report (agent-ready-codebase-attributes.md) defines these attributes.
All assessors must inherit from BaseAssessor and implement attribute_id and assess().

- |
We intentionally don't enforce line length (E501 ignored in ruff).
We use black for formatting with defaults.
Prefer explicit over clever, simple over complex.

- |
Error handling: Assessors return Finding.create_skipped() when tools missing, not crash.
Fail gracefully and provide actionable remediation guidance.

- |
Conventional commits (feat:, fix:, chore:, docs:, test:, refactor:).
All commits squashed before merge and signed with git signature.

- |
Target >80% test coverage on new code.
Tests focus on behavior, not implementation.
Mock external dependencies, don't over-mock internals.

- |
CLAUDE.md is source of truth for development.
README.md is user-facing.
All new features must update CLAUDE.md before merging.

# Professional, concise tone (default: "")
tone_instructions: |
- Be direct and concise - no fluff or pleasantries
- Focus on "why" not just "what" (explain the impact)
Expand Down
Loading