This repository was archived by the owner on Jan 2, 2026. It is now read-only.
feat: Observability instrumentation and distributed tracing#24
Merged
Conversation
- Create PROGRESS.md with 29 tasks across 6 phases - Update README.md with started timestamp - Spec approved and ready for implementation Implements: SPEC-2025-12-25-001 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
Complete observability foundation with: Core Modules: - config.py: ObservabilityConfig with env-based configuration - metrics.py: Thread-safe MetricsCollector (counters, histograms, gauges) - tracing.py: Span context with contextvars-based propagation - session.py: SessionIdentifier with privacy-preserving hashes - decorators.py: measure_duration decorator (sync/async support) - logging.py: StructuredLogger with JSON/text formatters Exporters: - prometheus.py: Prometheus text format (no external deps) - json_exporter.py: Full JSON export for metrics and traces Key Features: - Lazy imports via __getattr__ for hook performance (<5ms import) - Thread-safe operations with threading.Lock - Privacy-preserving SHA256 hashes (8 char truncated) - Histogram buckets aligned with hook timeouts (2s, 5s, 15s, 30s) - Rolling window for bounded memory in histograms Tests: - 115 new tests across 8 test modules - Coverage: 87.76% (above 80% threshold) - All quality gates passing Closes Phase 1 of SPEC-2025-12-25-001 (Observability Instrumentation) Ref: #10 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
Implement comprehensive observability for git-notes-memory plugin: Phase 1 - Core Infrastructure: - MetricsCollector with counters, histograms, gauges (thread-safe) - SpanContext and trace_operation for distributed tracing - StructuredLogger with JSON/text formatters - SessionIdentifier for multi-tenant distinguishability - measure_duration decorator and timed_context Phase 2 - Service Instrumentation: - Instrument CaptureService, RecallService, EmbeddingService - Instrument IndexService, GitOps, Hook Handlers - Fix silent failure points with explicit logging and metrics - Add timed_hook_execution context manager Phase 3 - Structured Logging: - Add *args support to StructuredLogger for backwards compatibility - Migrate all 5 hook handlers to get_logger() - Migrate 7 hook support modules to structured logging Phase 4 - CLI & Export: - /memory:metrics command with text/json/prometheus formats - /memory:traces command with filtering options - /memory:health command with --timing for latency percentiles - PrometheusExporter class for Prometheus text format Phases 5-6 (OpenTelemetry, Docker stack) skipped as optional Tier 3. Closes #10 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
MAXALL deep-clean remediation: - CRIT-001: Add exponential backoff with jitter to lock acquisition - Replaces fixed 100ms retry with 50ms→2s exponential backoff - Adds 0-10% jitter to prevent thundering herd - Reduces resource exhaustion risk under high concurrency - HIGH-004: Add composite index on (namespace, timestamp DESC) - Improves range query performance for get_by_namespace() - HIGH-014: Create docs/observability.md documentation - Metrics collection and Prometheus export - Tracing with span hierarchies - Structured logging configuration - Health checks and performance characteristics Code review artifacts in docs/code-review/2025/12/25/: - CODE_REVIEW.md: 90 findings from 9 specialist agents - REVIEW_SUMMARY.md: Executive summary - REMEDIATION_TASKS.md: Actionable checklist - REMEDIATION_REPORT.md: Final status with false positives All 1949 tests passing, 87.70% coverage. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
There was a problem hiding this comment.
Pull request overview
This PR implements comprehensive observability instrumentation for the git-notes-memory plugin, addressing Issue #10. The implementation includes metrics collection, distributed tracing, and structured logging with CLI commands for accessing telemetry data.
Key Changes:
- New
observability/module with metrics, tracing, logging, and session identification - Instrumentation of 5 services (CaptureService, RecallService, EmbeddingService, IndexService, GitOps) and 5 hook handlers
- Three new CLI commands:
/memory:metrics,/memory:traces,/memory:health - Silent failure tracking with explicit logging instead of suppressed exceptions
- Privacy-preserving session identification for multi-tenant distinguishability
Reviewed changes
Copilot reviewed 54 out of 54 changed files in this pull request and generated 8 comments.
Show a summary per file
| File | Description |
|---|---|
tests/test_observability/*.py |
Comprehensive test suite (273-299 lines each) covering metrics, tracing, logging, session, exporters, and decorators |
src/git_notes_memory/observability/*.py |
Core observability modules: config, metrics, tracing, session, logging, decorators, exporters |
src/git_notes_memory/capture.py |
Instrumented with metrics/tracing, exponential backoff with jitter for lock acquisition, silent failure tracking |
src/git_notes_memory/recall.py |
Added measure_duration decorators and trace operations for search paths |
src/git_notes_memory/index.py |
Composite index added (namespace, timestamp), silent failure tracking for index operations |
src/git_notes_memory/sync.py |
Silent failure logging for hash verification |
src/git_notes_memory/git_ops.py |
Git command metrics and tracing, silent failure tracking for fetch operations |
src/git_notes_memory/embedding.py |
Model load timing and embedding generation metrics |
src/git_notes_memory/hooks/*.py |
All hook handlers instrumented with timed_hook_execution, migrated to structured logging |
docs/spec/active/2025-12-25-observability-instrumentation/*.md |
Complete spec documentation: requirements, architecture, implementation plan, decisions, research, progress |
docs/observability.md |
User-facing documentation for metrics, tracing, and logging features |
Phase 6 implementation (tasks 6.1-6.2): - docker-compose.yml with OTEL Collector, Prometheus, Tempo, Loki, Grafana - OpenTelemetry Collector config with OTLP receivers and exporters - Prometheus scrape configuration for OTEL metrics - Tempo distributed tracing backend configuration - Loki log aggregation configuration - Grafana datasource provisioning (Prometheus, Tempo, Loki) - Grafana dashboard provisioning with auto-discovery - memory-operations.json dashboard (captures, searches, latency) - hook-performance.json dashboard (executions, timeouts, p95 latency) Exposed ports: - 3000: Grafana UI - 9090: Prometheus UI - 4317/4318: OTEL gRPC/HTTP - 3100: Loki - 3200: Tempo Usage: cd docker && docker-compose up -d 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
Addresses PR #24 review comments: - SECURITY: Moved inline Python code from commands/*.md to scripts/*.py to prevent command injection via $ARGUMENTS string interpolation - scripts/metrics.py: Safe argument parsing via argparse/sys.argv - scripts/traces.py: Safe argument parsing via argparse/sys.argv - scripts/health.py: Safe argument parsing via argparse/sys.argv - Updated commands/metrics.md, traces.md, health.md to call scripts - Fixed exporters/__init__.py: explicit imports instead of lazy loading to resolve Copilot undefined export warnings The previous implementation embedded $ARGUMENTS directly into Python string literals, allowing code injection via crafted arguments. The new scripts pass arguments safely via command line. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
Completes Phase 5 of the observability spec by adding OTLP HTTP export
capability so telemetry can be pushed to OpenTelemetry Collector and the
Docker Compose observability stack.
Changes:
- New `exporters/otlp.py` with OTLPExporter class using stdlib only
- Converts internal Span/metrics to OTLP JSON format
- Pushes to {endpoint}/v1/traces and {endpoint}/v1/metrics
- Configurable via MEMORY_PLUGIN_OTLP_ENDPOINT env var
- Updated Stop handler to flush telemetry at session end
- Updated observability.md with OTLP configuration section
- Added comprehensive tests for OTLP exporter
- pyproject.toml: added S310 per-file-ignore for OTLP file
The exporter uses urllib.request (stdlib) to avoid adding dependencies.
When MEMORY_PLUGIN_OTLP_ENDPOINT is set, traces and metrics are
automatically exported when the Claude Code session ends.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <[email protected]>
Merged origin/main which includes: - Secrets filtering with automatic detection and redaction - PII detection and redaction patterns - Security allowlist management - Audit logging for security events - New `/memory:scan-secrets`, `/memory:audit-log` commands Conflict resolution in capture.py: - Preserved observability trace_operation context managers - Integrated filter_warnings from secrets filtering into warnings list - Combined warnings using "; ".join() for multiple warning sources 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
zircote
added a commit
that referenced
this pull request
Dec 26, 2025
* feat: initialize observability instrumentation implementation - Create PROGRESS.md with 29 tasks across 6 phases - Update README.md with started timestamp - Spec approved and ready for implementation Implements: SPEC-2025-12-25-001 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat(observability): implement Phase 1 core infrastructure Complete observability foundation with: Core Modules: - config.py: ObservabilityConfig with env-based configuration - metrics.py: Thread-safe MetricsCollector (counters, histograms, gauges) - tracing.py: Span context with contextvars-based propagation - session.py: SessionIdentifier with privacy-preserving hashes - decorators.py: measure_duration decorator (sync/async support) - logging.py: StructuredLogger with JSON/text formatters Exporters: - prometheus.py: Prometheus text format (no external deps) - json_exporter.py: Full JSON export for metrics and traces Key Features: - Lazy imports via __getattr__ for hook performance (<5ms import) - Thread-safe operations with threading.Lock - Privacy-preserving SHA256 hashes (8 char truncated) - Histogram buckets aligned with hook timeouts (2s, 5s, 15s, 30s) - Rolling window for bounded memory in histograms Tests: - 115 new tests across 8 test modules - Coverage: 87.76% (above 80% threshold) - All quality gates passing Closes Phase 1 of SPEC-2025-12-25-001 (Observability Instrumentation) Ref: #10 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat: Complete observability instrumentation (Phases 1-4) Implement comprehensive observability for git-notes-memory plugin: Phase 1 - Core Infrastructure: - MetricsCollector with counters, histograms, gauges (thread-safe) - SpanContext and trace_operation for distributed tracing - StructuredLogger with JSON/text formatters - SessionIdentifier for multi-tenant distinguishability - measure_duration decorator and timed_context Phase 2 - Service Instrumentation: - Instrument CaptureService, RecallService, EmbeddingService - Instrument IndexService, GitOps, Hook Handlers - Fix silent failure points with explicit logging and metrics - Add timed_hook_execution context manager Phase 3 - Structured Logging: - Add *args support to StructuredLogger for backwards compatibility - Migrate all 5 hook handlers to get_logger() - Migrate 7 hook support modules to structured logging Phase 4 - CLI & Export: - /memory:metrics command with text/json/prometheus formats - /memory:traces command with filtering options - /memory:health command with --timing for latency percentiles - PrometheusExporter class for Prometheus text format Phases 5-6 (OpenTelemetry, Docker stack) skipped as optional Tier 3. Closes #10 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * fix: add exponential backoff and observability documentation MAXALL deep-clean remediation: - CRIT-001: Add exponential backoff with jitter to lock acquisition - Replaces fixed 100ms retry with 50ms→2s exponential backoff - Adds 0-10% jitter to prevent thundering herd - Reduces resource exhaustion risk under high concurrency - HIGH-004: Add composite index on (namespace, timestamp DESC) - Improves range query performance for get_by_namespace() - HIGH-014: Create docs/observability.md documentation - Metrics collection and Prometheus export - Tracing with span hierarchies - Structured logging configuration - Health checks and performance characteristics Code review artifacts in docs/code-review/2025/12/25/: - CODE_REVIEW.md: 90 findings from 9 specialist agents - REVIEW_SUMMARY.md: Executive summary - REMEDIATION_TASKS.md: Actionable checklist - REMEDIATION_REPORT.md: Final status with false positives All 1949 tests passing, 87.70% coverage. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat(observability): add Docker Compose local observability stack Phase 6 implementation (tasks 6.1-6.2): - docker-compose.yml with OTEL Collector, Prometheus, Tempo, Loki, Grafana - OpenTelemetry Collector config with OTLP receivers and exporters - Prometheus scrape configuration for OTEL metrics - Tempo distributed tracing backend configuration - Loki log aggregation configuration - Grafana datasource provisioning (Prometheus, Tempo, Loki) - Grafana dashboard provisioning with auto-discovery - memory-operations.json dashboard (captures, searches, latency) - hook-performance.json dashboard (executions, timeouts, p95 latency) Exposed ports: - 3000: Grafana UI - 9090: Prometheus UI - 4317/4318: OTEL gRPC/HTTP - 3100: Loki - 3200: Tempo Usage: cd docker && docker-compose up -d 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * fix(security): divest inline Python from command files into scripts Addresses PR #24 review comments: - SECURITY: Moved inline Python code from commands/*.md to scripts/*.py to prevent command injection via $ARGUMENTS string interpolation - scripts/metrics.py: Safe argument parsing via argparse/sys.argv - scripts/traces.py: Safe argument parsing via argparse/sys.argv - scripts/health.py: Safe argument parsing via argparse/sys.argv - Updated commands/metrics.md, traces.md, health.md to call scripts - Fixed exporters/__init__.py: explicit imports instead of lazy loading to resolve Copilot undefined export warnings The previous implementation embedded $ARGUMENTS directly into Python string literals, allowing code injection via crafted arguments. The new scripts pass arguments safely via command line. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat(observability): implement OTLP HTTP exporter for telemetry push Completes Phase 5 of the observability spec by adding OTLP HTTP export capability so telemetry can be pushed to OpenTelemetry Collector and the Docker Compose observability stack. Changes: - New `exporters/otlp.py` with OTLPExporter class using stdlib only - Converts internal Span/metrics to OTLP JSON format - Pushes to {endpoint}/v1/traces and {endpoint}/v1/metrics - Configurable via MEMORY_PLUGIN_OTLP_ENDPOINT env var - Updated Stop handler to flush telemetry at session end - Updated observability.md with OTLP configuration section - Added comprehensive tests for OTLP exporter - pyproject.toml: added S310 per-file-ignore for OTLP file The exporter uses urllib.request (stdlib) to avoid adding dependencies. When MEMORY_PLUGIN_OTLP_ENDPOINT is set, traces and metrics are automatically exported when the Claude Code session ends. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> --------- Co-authored-by: Claude Opus 4.5 <[email protected]>
zircote
added a commit
that referenced
this pull request
Dec 26, 2025
* feat: initialize observability instrumentation implementation - Create PROGRESS.md with 29 tasks across 6 phases - Update README.md with started timestamp - Spec approved and ready for implementation Implements: SPEC-2025-12-25-001 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat(observability): implement Phase 1 core infrastructure Complete observability foundation with: Core Modules: - config.py: ObservabilityConfig with env-based configuration - metrics.py: Thread-safe MetricsCollector (counters, histograms, gauges) - tracing.py: Span context with contextvars-based propagation - session.py: SessionIdentifier with privacy-preserving hashes - decorators.py: measure_duration decorator (sync/async support) - logging.py: StructuredLogger with JSON/text formatters Exporters: - prometheus.py: Prometheus text format (no external deps) - json_exporter.py: Full JSON export for metrics and traces Key Features: - Lazy imports via __getattr__ for hook performance (<5ms import) - Thread-safe operations with threading.Lock - Privacy-preserving SHA256 hashes (8 char truncated) - Histogram buckets aligned with hook timeouts (2s, 5s, 15s, 30s) - Rolling window for bounded memory in histograms Tests: - 115 new tests across 8 test modules - Coverage: 87.76% (above 80% threshold) - All quality gates passing Closes Phase 1 of SPEC-2025-12-25-001 (Observability Instrumentation) Ref: #10 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat: Complete observability instrumentation (Phases 1-4) Implement comprehensive observability for git-notes-memory plugin: Phase 1 - Core Infrastructure: - MetricsCollector with counters, histograms, gauges (thread-safe) - SpanContext and trace_operation for distributed tracing - StructuredLogger with JSON/text formatters - SessionIdentifier for multi-tenant distinguishability - measure_duration decorator and timed_context Phase 2 - Service Instrumentation: - Instrument CaptureService, RecallService, EmbeddingService - Instrument IndexService, GitOps, Hook Handlers - Fix silent failure points with explicit logging and metrics - Add timed_hook_execution context manager Phase 3 - Structured Logging: - Add *args support to StructuredLogger for backwards compatibility - Migrate all 5 hook handlers to get_logger() - Migrate 7 hook support modules to structured logging Phase 4 - CLI & Export: - /memory:metrics command with text/json/prometheus formats - /memory:traces command with filtering options - /memory:health command with --timing for latency percentiles - PrometheusExporter class for Prometheus text format Phases 5-6 (OpenTelemetry, Docker stack) skipped as optional Tier 3. Closes #10 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * fix: add exponential backoff and observability documentation MAXALL deep-clean remediation: - CRIT-001: Add exponential backoff with jitter to lock acquisition - Replaces fixed 100ms retry with 50ms→2s exponential backoff - Adds 0-10% jitter to prevent thundering herd - Reduces resource exhaustion risk under high concurrency - HIGH-004: Add composite index on (namespace, timestamp DESC) - Improves range query performance for get_by_namespace() - HIGH-014: Create docs/observability.md documentation - Metrics collection and Prometheus export - Tracing with span hierarchies - Structured logging configuration - Health checks and performance characteristics Code review artifacts in docs/code-review/2025/12/25/: - CODE_REVIEW.md: 90 findings from 9 specialist agents - REVIEW_SUMMARY.md: Executive summary - REMEDIATION_TASKS.md: Actionable checklist - REMEDIATION_REPORT.md: Final status with false positives All 1949 tests passing, 87.70% coverage. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat(observability): add Docker Compose local observability stack Phase 6 implementation (tasks 6.1-6.2): - docker-compose.yml with OTEL Collector, Prometheus, Tempo, Loki, Grafana - OpenTelemetry Collector config with OTLP receivers and exporters - Prometheus scrape configuration for OTEL metrics - Tempo distributed tracing backend configuration - Loki log aggregation configuration - Grafana datasource provisioning (Prometheus, Tempo, Loki) - Grafana dashboard provisioning with auto-discovery - memory-operations.json dashboard (captures, searches, latency) - hook-performance.json dashboard (executions, timeouts, p95 latency) Exposed ports: - 3000: Grafana UI - 9090: Prometheus UI - 4317/4318: OTEL gRPC/HTTP - 3100: Loki - 3200: Tempo Usage: cd docker && docker-compose up -d 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * fix(security): divest inline Python from command files into scripts Addresses PR #24 review comments: - SECURITY: Moved inline Python code from commands/*.md to scripts/*.py to prevent command injection via $ARGUMENTS string interpolation - scripts/metrics.py: Safe argument parsing via argparse/sys.argv - scripts/traces.py: Safe argument parsing via argparse/sys.argv - scripts/health.py: Safe argument parsing via argparse/sys.argv - Updated commands/metrics.md, traces.md, health.md to call scripts - Fixed exporters/__init__.py: explicit imports instead of lazy loading to resolve Copilot undefined export warnings The previous implementation embedded $ARGUMENTS directly into Python string literals, allowing code injection via crafted arguments. The new scripts pass arguments safely via command line. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> * feat(observability): implement OTLP HTTP exporter for telemetry push Completes Phase 5 of the observability spec by adding OTLP HTTP export capability so telemetry can be pushed to OpenTelemetry Collector and the Docker Compose observability stack. Changes: - New `exporters/otlp.py` with OTLPExporter class using stdlib only - Converts internal Span/metrics to OTLP JSON format - Pushes to {endpoint}/v1/traces and {endpoint}/v1/metrics - Configurable via MEMORY_PLUGIN_OTLP_ENDPOINT env var - Updated Stop handler to flush telemetry at session end - Updated observability.md with OTLP configuration section - Added comprehensive tests for OTLP exporter - pyproject.toml: added S310 per-file-ignore for OTLP file The exporter uses urllib.request (stdlib) to avoid adding dependencies. When MEMORY_PLUGIN_OTLP_ENDPOINT is set, traces and metrics are automatically exported when the Claude Code session ends. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]> --------- Co-authored-by: Claude Opus 4.5 <[email protected]>
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 subscribe to this conversation on GitHub.
Already have an account?
Sign in.
1 participant
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.
Summary
Implements comprehensive observability instrumentation for git-notes-memory (Issue #10, SPEC-2025-12-25-001).
Phase 1: Core Infrastructure ✅
observability/module with lazy imports for hook performanceObservabilityConfigwith environment-based configurationMetricsCollector(counters, histograms, gauges)SpanContextwith contextvars-based trace propagationSessionIdentifierwith privacy-preserving SHA256 hashesmeasure_durationdecorator (sync/async support)Phase 2: Service Instrumentation ✅
timed_hook_executioncontext manager for hook handlersPhase 3: Structured Logging ✅
StructuredLoggerwith JSON/text formattersPhase 4: CLI & Export ✅
/memory:metricscommand (--format=text|json|prometheus)/memory:tracescommand (--operation,--status,--limit)/memory:health --timingfor latency percentilesPhase 5: OpenTelemetry (Skipped)
Phase 6: Local Stack & Dashboards ✅
memory-operations.json- Captures, searches, latency percentileshook-performance.json- Executions, timeouts, p95 latency by hookCode Review Fixes ✅
Test Plan
Files Changed
src/git_notes_memory/observability/- New observability module (8 files)src/git_notes_memory/*.py- Instrumented core servicessrc/git_notes_memory/hooks/*.py- Migrated to structured loggingcommands/metrics.md,traces.md,health.md- New CLI commandsdocker/- Docker Compose observability stack (9 files)docs/observability.md- Observability documentationdocs/code-review/- Code review artifacts🤖 Generated with Claude Code