fix: return 'Unknown' as primary_language when directory has no Python files

fix: move success computation to end of run_tests and remove it from except block

fix: use sys.executable -m pytest to ensure pytest is always findable in sandbox

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

fix: suppress bandit B404 warning on subprocess import in pytest_runner

fix: import subprocess in pytest_runner so timeout patch target exists and handle TimeoutExpired

fix: import subprocess in pytest_runner so timeout patch target exists and handle TimeoutExpired

fix

fix: suppress bandit B110 and B404 warnings in file_scanner and sandbox

fix: correct sandbox process cast and scan_repository return type to dict

fix: correct sandbox process cast and scan_repository return type to dict

fix: correct sandbox process cast and scan_repository return type to dict

fix: resolve mypy errors in sandbox.py and file_scanner.py

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

security: harden prompt injection defense with XML delimiters and injection pattern stripping

security: add sandboxed subprocess runner with CPU/memory resource limits

feat(security): add Semgrep scanner with python + secrets rulesets

feat: add security depth — secret scanning, dependency vulns, report surface

fix: complete Redis cache integration and clean up test layer

fix

fix: add missing file_list param to read_source_samples and wire changed_files through

fix: feat: fault-tolerant parallel agents, AST security scanner, and shared source reader

fixed lint

style(github_tool): fix import grouping, remove trailing whitespace

fix(security): resolve all bandit findings — log swallowed exception, annotate justified subprocess usage

feat(tools): add diff-aware file filtering for incremental reviews

fix(types): wrap Repo.clone_from in lambda to fix mypy arg-type errors

fix(github_tool): add 60s timeout to Repo.clone_from to prevent worker starvation

security: sanitize file content before LLM injection to mitigate prompt injection attacks

fix(github_tool): cleanup partial clone directory on Repo.clone_from failure

fix: use sys.executable -m pytest in pytest_runner for venv compatibility

fix: log warnings instead of silently swallowing file read errors in agents and scanner

fix: validate GitHub URL inside clone_repository to protect all callers not just API layer

fix: guard against ValueError when parsing pytest output integers

fix

fix(security): shallow clone (depth=1) to limit bandwidth before size check

fix: remove eager wildcard imports from __init__.py files

fix: enforce max_repo_size_mb after clone

fix: resolve all ruff lint errors (E402, F401, F811)

fix: use unique per-request clone paths to prevent concurrent collisions

refactor: replace print() with structured logging across all agents and tools

fix: use Windows-compatible temp path for repository cloning

fix: suppress remaining bandit warnings with nosec comments

fix: add nosec comments and fix bandit txt format in static analyzer

fix: resolve all ruff linting errors in __init__ files and CI

feat: add all tools - github, scanner, ast, pytest, static analyzer

chore: initialize project folder structure