Adds a management command and helper script to configure the Jumbo API external data source and fetch chat sessions.
Ensures idempotent creation and linkage to the Jumbo company while reading API credentials from environment variables.
Prints data source details and a post-fetch summary to ease setup and verification.
Removes an unused assertion import and aliases the unused test parameter to suppress linter warnings.
Keeps the test scaffold clean without changing behavior.
Introduces uv-based Docker workflow with non-root runtime, cached installs, and uv-run for web and Celery. Updates docker-compose to Postgres + Redis, loads .env, and removes source bind mount for reproducible builds.
Switches settings to use Postgres when env is present with SQLite fallback; broadens allowed hosts for containerized development. Adds psycopg2-binary and updates sample env for Redis in Docker.
Adds company scoping to external data models and links sessions during ingestion; provides management commands to seed a Jumbo company/users and sync external chat data into the dashboard.
Includes .dockerignore, TypeScript config and typings, and minor template/docs tweaks.
Requires database migration.
Introduces Playwright testing agents (planner, generator, healer) powered by MCP to plan, generate, and heal end-to-end tests.
Configures the MCP server and integrates agent workflows into OpenCode and GitHub chat modes to enable AI-assisted testing.
Adds Playwright test dependency and updates lockfile; adjusts markdown lint ignores to reduce noise.
Adds contributor guidance for Claude Code to streamline local development.
Normalizes shell script shebangs to use /usr/bin/env bash for portability.
Enables automated browser testing workflows and resilient test maintenance within AI-enabled tooling.
- Skip prettier-jinja and prettier-all (require bunx not available in CI)
- Skip tombi-format and tombi-lint (network issues fetching schema catalog)
- Hooks still run locally where dependencies are available
- Resolves pre-commit.ci failures
- Add top-level 'multi-ecosystem-groups' section
- Move schedule to group level for single consolidated schedule
- Add 'patterns: ["*"]' at ecosystem level (required)
- All ecosystems assigned to 'all-dependencies' group
- Results in single monthly PR with updates across all ecosystems
- Replace 'multi-ecosystem-group' with proper 'groups' syntax
- Add required 'patterns' property with wildcard matcher
- Group all dependencies per ecosystem for cleaner PRs
- Fixes dependabot configuration validation error
Points the config to the correct CLI2 schema and removes the outdated nested schema key.
Improves editor validation and IntelliSense by aligning with the current tool schema.
Enables automated dependency updates across uv, bun, GitHub Actions, Docker, Compose, and devcontainers
Groups updates across ecosystems to reduce PR noise
Schedules monthly runs to balance freshness and maintenance effort
- Wrap config in 'config' key for markdownlint-cli2
- Use MD013 rule name instead of 'line-length' alias
- Disable MD013 line-length checks
- Add allowed languages: sh, python, csv, tree
- Fix broken link reference in TODO.md
- All markdown linting now passes (37 errors -> 0)
Configure oxc LSP initialization options: run set to "onType", typeAware
enabled, unusedDisableDirectives set to "allow", and
configPath/tsConfigPath left null.
- Move setupAjaxPagination to outer scope in ajax-pagination.js
- Move setupAjaxNavigation, reloadScripts, initializePageScripts to outer scope in ajax-navigation.js
- Move updatePlotlyTheme, resizeCharts, updateDashboardStats, updateDashboardCharts to outer scope in dashboard.js
- Move handleSidebarOnResize, setTheme, getSystemPreference to outer scope in main.js
- Avoid recreating functions on every DOMContentLoaded call
- All oxlint strict checks now pass (11 warnings -> 0)
- Fix typescript-eslint(no-implied-eval) warning
- Use arrow function instead of direct function reference
- All oxlint type-aware checks now pass (1 warning -> 0)
- Add ty.toml configuration with Django project root
- Add py.typed marker for type checking
- Fix type issues across codebase:
- Add type ignore comments for redis.exceptions imports
- Fix django.db.models.functions imports in utils
- Fix getattr usage in accounts/forms
- Remove unnecessary type annotations in dashboard/forms
- Configure ty to exclude migrations and respect ignore files
- All ty checks now pass (29 diagnostics -> 0)