fix: patch AsyncResult in job poll test to simulate completed task

fix: add missing 'success' key to run_tests return dict

fix: add missing 'success' key to run_tests return dict

fix: loosen timeout message assertion to match sandbox output format

fix: correct all async route status code assertions (POST=202, GET poll=200)

fix: correct expected status code in test_async_analyze_returns_job_id to 202

fix: set mock_result.status and correct status code assertion in poll success test

fix: correct expected status code in job poll test to 200

fix

fix

fix: correct expected status code in async job submission test to 202

infra(terraform): add EKS cluster, Elasticache Redis, and ECR modules for cloud deployment

eval: expand benchmark dataset from 6 to 15 fixtures covering race conditions, TOCTOU, async misuse, N+1, path traversal, insecure deserialization, and missing auth

eval: fix filename typo and add real-model eval suite gated by RUN_REAL_EVAL env var

test(integration): add end-to-end test for async job submission and poll-to-completion cycle

test(load): add Locust load test suite for /analyze and /analyze/async endpoints

feat: add integration tests, GitHub API mocking, and Locust load tests

feat: add evaluation framework with precision/recall metrics and model comparison

fix: complete Redis cache integration and clean up test layer

fix

feat: add Redis/Celery async job queue for repository analysis

fix

fix

fix

fix: update mock target for get_python_files in test after refactor to file_scanner

test: resolve failing llm token usage tests by fixing patch namespace

test: fix failing llm token usage tests caused by wrong patch namespace

fixed lint

test(metrics/ast_parser/pytest_runner): cover error counters, cost estimation, syntax errors, timeout paths

test(code_review_agent): cover incremental diff path, test-file skipping, low-score second pass, file read error

test(routes): cover 500 error paths, cleanup on failure, feedback endpoints, limiter key

test(github_tool): cover clone success/timeout/error/size-limit, get_changed_files, delete

test(llm): cover HTTP call, token recording, rate-limit helper, JSON parse fallback

fix(tests): add specialist_review_agent mock responses to full pipeline test

fix(tests): respect call_llm_for_json return type in full pipeline mock

fix(tests): update test_full_pipeline to patch get_settings(), remove unused json import, fix LLM_RESPONSES dict→str

fix(tests): update test_api_routes to patch get_settings() instead of removed module-level settings

fix(tests): serialize code_review LLM response to JSON string — raw dict caused TypeError in call_llm_for_json

fix(tests): remove phantom nested function in test_repo_analysis_agent — assertions were never executing

fix(tests): replace hardcoded /tmp paths with pytest tmp_path fixture for cross-platform compatibility

fix(tests): patch call_llm_for_json in code_review_agent; return dict in LLM_RESPONSES

fix(tests): patch call_llm_for_json in bug_detection_agent test; return dict not string

fix(tests): patch call_llm_for_json (not call_llm) in repo_analysis and bug_detection agent tests

fix(tests): update pipeline test patches after call_llm_for_json refactor

fix(tests): patch call_llm at each agent import site, not at definition module

test: add full pipeline integration test covering HTTP → agents → response validation

fix: disable slowapi rate limiter in tests to prevent IP extraction crash with TestClient

fix: add clone_repository mock to auth rejection tests to prevent git crash before 401

fix: mock clone_repository in auth rejection tests to prevent git crash before 401 is returned

fix: remove redundant local patch imports, rely on module-level import in test_agents.py