- Updated PM2 version to 6.0.14 in package.json and pnpm-lock.yaml for improved stability.
- Changed interpreter for PM2 processes to use `process.execPath` for consistency.
- Added a new `start:pm2` script in package.json to streamline PM2 process management.
- Updated README to reflect changes in PM2 usage and instructions for starting the application.
- Renamed the control process to `jflow-control` and updated the script path to `control.js`.
- Introduced a new `jflow-web` process for the production UI, serving on port 8500.
- Enhanced environment variable handling for both processes, ensuring proper port assignments.
- Updated README to reflect new process architecture and usage instructions for starting the application.
- Updated AGENTS.md to include a `name` field for scripts, allowing custom display titles in the editor and graph.
- Enhanced script icon resolution to support `.jpeg` file extensions.
- Modified workflow parsing to accommodate the new `name` property for steps, improving step identification.
- Added new utility functions for handling custom step names and display names.
- Introduced tests to validate the functionality of custom step names in workflow YAML parsing and step labeling.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Added a new `reset-admin` script to reset or create the admin username and password.
- Updated the README to include instructions for resetting the admin login.
- Enhanced user management by allowing username updates during admin reset.
- Defaulted new resources to the internal namespace `local` for better organization.
- Removed unused owner selection from various components to streamline the UI.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Renamed `ensureDataObject` to `passContext` and added `mergeData` for better context management.
- Updated `detectUrlChanges` to utilize the new context handling functions, improving data merging and response handling.
- Enhanced evaluation of JSONata expressions with a new `evalCtx` function to streamline context passing.
- Adjusted test assertions to reflect changes in output structure from `data` to `output`.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Migrated legacy workflow files from `packages/server/workflows/` to `packages/server/data/workflows/` when the new store is empty.
- Introduced new API endpoints to list and retrieve example workflows.
- Added a dialog component for selecting example workflows when creating new workflows.
- Updated workflow paths and configurations to reflect the new structure.
- Enhanced the README and AGENTS.md documentation to clarify workflow management and usage.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Added `workflowLastModifiedAt` function to retrieve the last modified timestamp of workflow YAML files.
- Introduced `encodeBinaryForWire` and `reviveBinaryFromWire` functions for better handling of binary data in JSON.
- Implemented `useDeleteKv` hook for deleting key-value pairs in the web API.
- Enhanced `scriptsPluginFactory` to support dry-run evaluations with JSONata expressions.
- Added `set-dry-run-smoke.js` test to validate dry-run functionality.
- Updated profile configuration to include `overlayFromMerged` for better profile management.
- Introduced try-session management in the web components to handle step execution states.
- Improved various components to support new try-session features and maintain UI consistency.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
Keep only example workflows in default/, move site-specific YAMLs to gitignored local/, remove the duplicate get-current-time plugin install, document personal plugins, clarify PM2 ecosystem usage, add shared Vitest and web ESLint, and extract OpsStatusCard for ops cards.
Co-authored-by: Cursor <cursoragent@cursor.com>
Consolidate format helpers under lib/format, split React Query hooks by domain, extract AuthPicker/ConfigRefHint and HTTP trigger routing, demote file-private exports, and move KV/secrets/variables/profiles/http-auths stores under src/stores with root re-exports.
Co-authored-by: Cursor <cursoragent@cursor.com>
Share profile merge, workflow filename/path helpers, HTTP_METHODS (with OPTIONS), and isPlainObject so web and server stop drifting. Core script sandboxes keep local isPlainObject copies.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Updated package manager to pnpm@11.22.0 in package.json.
- Removed unused dependencies and scripts from package.json and pnpm-lock.yaml.
- Cleaned up fs-store.js and other files by removing commented-out code and unused imports.
- Updated workflow configurations to use environment variables for URLs.
Profiles store script + default config per owner. Workflow steps reference
them with profile:, overlay keys win, and the UI marks overrides. Profile
name is immutable after create so YAML refs stay stable.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
Extend GET /api/runs with offset/total, date and trigger filters,
and configurable column sort. Rebuild Events page with URL-driven state
matching the KV page pagination pattern.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Introduced `disableOnConsecutiveFailures` option in workflow triggers to automatically disable workflows after reaching a specified failure threshold.
- Updated related functions to handle the new feature, including persistence of the disabled state and reloading of registries.
- Enhanced UI components to support the new option, allowing users to toggle the auto-disable feature in the workflow configuration.
Mount auth routes on the control plane and proxy /api/auth to :8600 in
dev:pm2 so the UI can authenticate while the HTTP API process is down.
Co-authored-by: Cursor <cursoragent@cursor.com>
Import getConfigGeneration, startHeartbeatLoop, and subscribeReload so
the monolith runner starts without ReferenceError crashes.
Co-authored-by: Cursor <cursoragent@cursor.com>
Restore the missing export so workflow duplication resolves YAML stems
without a broken import from workflow-duplicate.js.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Store up to 50 revisions per workflow in SQLite with normalized SHA dedup
- Soft-delete workflows to trash (7-day retention) with restore and permanent purge
- Assign UUID filenames for new and duplicated workflows; show name + file in UI
- Warn on invalid YAML, unknown scripts, and plaintext secrets with save-anyway option
- Add workflow backup zip (workflows + plugins) and merge/replace restore
- Trash page, history panel on editor, and smoke test
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
Load user plugins from /plugins instead of data/plugins so site-specific scripts stay in git and workflows can reference plugin/joplin-api and plugin/send-sms.
Co-authored-by: Cursor <cursoragent@cursor.com>
Introduce plugin/<id> installs (zip, HTTPS git, example, fork),
jerapah-plugin.json manifests with jerapah semver ranges, isolated
plugin dirs under data/plugins, and app version 0.1.0. Core scripts
are non-editable; get-current-time moves to examples/plugins.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
Vite listens on 8500 and proxies /ops to the control plane. Admin Ops UI
covers pause/resume, scale, drain/force restart, and restart-needed banner.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
Wire optional REDIS_PASS into the BullMQ ioredis connection, redact
credentials in startup logs, and document REDIS_PASS plus JFLOW_ROLE.
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Introduced a new SFTP test workflow in `test-sftp.yaml` to list remote files using SFTP protocol.
- Updated `registers.yaml` to include the new SFTP test workflow.
Enqueue HTTP, cron, and manual runs via Redis/BullMQ; workers execute
asynchronously with configurable concurrency. Triggers return 202 and
clients poll GET /api/runs/:id for progress (queued → running → done).
Co-authored-by: Nasyarobby Putra <nasyarobby@gmail.com>
- Added a new function to list consecutive failure streaks for workflows, allowing tracking of workflows that have failed multiple times in a row.
- Integrated the consecutive failure data into the dashboard, displaying streaks and counts for better visibility of workflow health.
- Created a dedicated FailuresPage to present detailed information about consecutive failures, enhancing user experience and monitoring capabilities.
- Updated API endpoints and hooks to support the new failure tracking features, ensuring seamless data retrieval and display.
This enhancement improves the ability to monitor and respond to workflow failures, contributing to overall system reliability.
- Updated all instances of `triggerWorkflow` to `onFailureWorkflow` across the codebase to improve clarity and consistency in naming.
- Adjusted related logic in workflow configurations and UI components to reflect the new naming convention.
- Enhanced error messages and documentation to align with the updated terminology.
This change aims to provide a clearer understanding of the workflow's failure handling mechanism.
nsrb
