codeninja 2.0.0

package/tasks/ask-route-path.task.md

@@ -0,0 +1,29 @@
+---
+type: task
+name: ask-route-path
+---
+
+Read `context.current_api.module_name` to suggest a default path.
+Suggested default: /[module_name_lowercase]
+
+Ask the user exactly this question:
+
+"Enter the route path for this endpoint."
+
+Show suggested default: /[module_name_lowercase]
+
+Examples:
+  /products
+  /products/:id
+  /orders/:orderId/items
+
+Wait for user input.
+
+Validate:
+- Must start with /
+- Use :paramName for URL parameters
+- Check against `context.api_routes` — must not be an exact duplicate of method+path in same service
+
+Store result in: `context.current_api.route_path`
+
+Do not ask any other question in this task.

package/tasks/ask-seed-row-values.task.md

@@ -0,0 +1,42 @@
+---
+type: task
+name: ask-seed-row-values
+---
+
+Read `context.current_db.columns` to know what columns exist.
+Read `context.current_db.seed_rows` to know which row number we are on.
+
+Let current_row = length of seed_rows + 1.
+Let total_rows = context.current_db.seed_count.
+
+Display: "Entering values for row [current_row] of [total_rows]"
+
+For each column in `context.current_db.columns`
+(SKIP: `id` — auto-generated, `created_at` — auto-set):
+
+  Ask the user exactly this question:
+  "[col_name] ([type]):"
+
+  Show additional context if available:
+  - If column has CHECK constraint → show: "Allowed: [values]"
+  - If column has DEFAULT → show: "Default: [default] (press enter to use)"
+  - If column is `password` → warn: "⚠ Enter encrypted/hashed value only. Never plaintext."
+  - If column is BOOLEAN → show: "Enter: true or false"
+  - If column is a NUMERIC → show expected format: "Example: 0.00000000"
+
+  Wait for user input.
+
+  If user presses enter on a column with a DEFAULT → use the DEFAULT value.
+
+  Validate:
+  - CHECK constraint values → must match allowed list
+  - BOOLEAN → must be true/false
+  - INTEGER → must be numeric
+  - NUMERIC → must match precision format
+
+  Store: value for this column in current row
+
+After all columns collected for this row:
+  Append complete row object to `context.current_db.seed_rows[]`
+
+Do not ask any other question in this task beyond column value collection.

package/tasks/ask-seed-rows-count.task.md

@@ -0,0 +1,22 @@
+---
+type: task
+name: ask-seed-rows-count
+---
+
+Read `context.current_db.table_name` to show context.
+
+Ask the user exactly this question:
+
+"How many rows do you want to seed into '[table_name]'?"
+
+Wait for user input.
+
+Validate:
+- Must be a positive integer
+- Maximum 500 rows per seed operation (for performance)
+- If more than 500 → inform: "For large datasets, consider importing a CSV
+  directly into the database. Seed files work best for small reference data."
+
+Store result in: `context.current_db.seed_count`
+
+Do not ask any other question in this task.

package/tasks/ask-service-description.task.md

@@ -0,0 +1,16 @@
+---
+type: task
+name: ask-service-description
+---
+
+Ask the user exactly this question:
+
+"Enter a short description for this service."
+
+Example: "Handles authentication and authorization"
+
+Wait for user input.
+
+Store result in: `context.current_init.description`
+
+Do not ask any other question in this task.

package/tasks/ask-service-name.task.md

@@ -0,0 +1,27 @@
+---
+type: task
+name: ask-service-name
+---
+
+Read `context.services` to get list of already initialized services.
+
+If services exist, show them to the user:
+"Existing services: [list]"
+
+Ask the user exactly this question:
+
+"Enter a name for this service."
+
+Example: auth, ledger, payments, notifications
+
+Wait for user input.
+
+Validate:
+- Lowercase only
+- No spaces (hyphens or underscores allowed)
+- Must NOT already exist in `context.services`
+- If duplicate → inform user and ask again
+
+Store result in: `context.current_init.service_name`
+
+Do not ask any other question in this task.

package/tasks/ask-service-port.task.md

@@ -0,0 +1,24 @@
+---
+type: task
+name: ask-service-port
+---
+
+Read `context.services` and collect all ports already in use.
+
+If ports are in use, show them:
+"Ports already in use: [list]"
+
+Ask the user exactly this question:
+
+"Enter the port number for this service."
+
+Wait for user input.
+
+Validate:
+- Must be a valid port number (1024–65535)
+- Must NOT conflict with any port already in `context.services`
+- If conflict → inform user and ask again
+
+Store result in: `context.current_init.port`
+
+Do not ask any other question in this task.

package/tasks/ask-supported-languages.task.md

@@ -0,0 +1,40 @@
+---
+type: task
+name: ask-supported-languages
+---
+
+Ask the user exactly this question:
+
+"Which languages should this service support?"
+
+Show guidance:
+"Enter language codes separated by commas. A separate language file will
+be created for each one under languages/. The first language in the list
+will be used as the default fallback."
+
+Example input: en
+Example input: en, ar
+Example input: en, ar, fr, de
+
+Wait for user input.
+
+Parse input:
+- Split by comma
+- Trim whitespace from each code
+- Convert to lowercase
+- Remove duplicates
+
+Validate:
+- Must have at least one language
+- Each code should be 2-5 characters (ISO 639-1 or locale codes)
+- If user enters something like "english" → inform them to use
+  the language code format (e.g. "en") and ask again
+
+Store result in: `context.current_init.supported_languages`
+as an array: ["en", "ar", "fr"]
+
+Inform user:
+"Default language will be: [first_in_list]
+Language files to be created: [en.js, ar.js, ...]"
+
+Do not ask any other question in this task.

package/tasks/ask-table-file-number.task.md

@@ -0,0 +1,36 @@
+---
+type: task
+name: ask-table-file-number
+---
+
+Scan the `database/<db_type>/migrations/` directory (from context.db.type).
+
+Read all existing filenames and extract their numeric prefixes.
+Find the highest existing number.
+
+Suggest: highest_number + 1
+
+Special rule: if the file being created is a shared index file, always suggest 111
+(or the next available number above 100 if 111 is taken).
+
+Display to user:
+
+"Existing migration files: [list filenames in order]"
+"Next available number: [suggested_number]"
+
+Ask the user exactly this question:
+
+"Enter the file number for this migration."
+
+Show suggested default: [suggested_number]
+
+If user presses enter without input → use suggested value.
+
+Validate:
+- Must be a positive integer
+- Must not already be in use by an existing file
+- If conflict → inform user and ask again
+
+Store result in: `context.current_db.file_number`
+
+Do not ask any other question in this task.

package/tasks/ask-table-indexes.task.md

@@ -0,0 +1,47 @@
+---
+type: task
+name: ask-table-indexes
+---
+
+Read `context.current_db.columns` to analyze the columns defined so far.
+Read `context.current_db.needs_status` and `context.current_db.needs_soft_delete`.
+
+Agent auto-generates suggested indexes based on rules from database-agent.md:
+
+1. For every column ending in `_id` → suggest: `idx_<table>_<col_name>`
+2. If both `status` and `is_deleted` exist → suggest compound: `idx_<table>_status_is_deleted`
+3. If `created_at` exists on an event/log table → suggest: `idx_<table>_created_at` with DESC
+4. If `email` exists → suggest: `idx_<table>_email`
+5. If `symbol` exists → suggest: `idx_<table>_symbol`
+
+Display to user:
+
+"Suggested indexes based on your columns:
+  ✓ idx_<table>_<col1>   ON (<col1>)
+  ✓ idx_<table>_<col2>   ON (<col2>)
+  ✓ idx_<table>_status_is_deleted  ON (status, is_deleted)"
+
+Ask the user exactly this question:
+
+"Accept all suggested indexes?"
+
+Present options:
+1. Yes — use all suggested indexes
+2. Remove some — I want to deselect specific ones
+3. Add more — I want to add additional indexes
+4. Start from scratch — define indexes manually
+
+Wait for user selection.
+
+If "Remove some":
+  List indexes with checkboxes
+  Ask which to remove — one at a time
+
+If "Add more":
+  Ask: "Describe the additional index (which column or columns?)"
+  Repeat until user is done
+
+Store final list in: `context.current_db.indexes[]`
+Each entry: `{ name: "idx_...", columns: [...], desc: true/false, where: null }`
+
+Do not ask any other question in this task beyond index selection.

package/tasks/ask-table-name.task.md

@@ -0,0 +1,32 @@
+---
+type: task
+name: ask-table-name
+---
+
+Read `context.db.schema.tables` to get list of available tables.
+
+If tables exist in context:
+  Ask the user exactly this question:
+
+  "Which table are you working with?"
+
+  List all tables from context.db.schema.tables.
+
+  Wait for user selection.
+
+If no tables exist:
+  Ask the user exactly this question:
+
+  "Enter the table name."
+
+  Wait for user input.
+  Validate: snake_case, plural form recommended (e.g. users, products, orders)
+
+If called from a @refactor workflow → store in: `context.current_refactor.table_name`
+If called from any @db:* workflow → store in: `context.current_db.table_name`
+
+Agent determines caller from `context.last_command`:
+- "refactor" → use current_refactor
+- any other value → use current_db
+
+Do not ask any other question in this task.

package/tasks/ask-table-needs-soft-delete.task.md

@@ -0,0 +1,29 @@
+---
+type: task
+name: ask-table-needs-soft-delete
+---
+
+Read `context.current_db.needs_status` — if true, auto-suggest YES for soft delete.
+
+Agent determines recommendation:
+- If table has status column → strongly recommend YES (soft delete pairs with status)
+- Event/log tables → recommend NO (events are immutable, never "deleted")
+- Reference data tables → depends on whether admin can deactivate records
+
+Ask the user exactly this question:
+
+"Should this table support soft delete? (is_deleted column instead of hard DELETE)"
+
+Show recommendation: "Suggested: [YES/NO] — [reason]"
+
+Present options:
+1. Yes — add is_deleted BOOLEAN NOT NULL DEFAULT FALSE
+2. No — records in this table will be hard deleted or never deleted
+
+Wait for user selection.
+
+Store result in: `context.current_db.needs_soft_delete`
+- Option 1 → true
+- Option 2 → false
+
+Do not ask any other question in this task.

package/tasks/ask-table-needs-status.task.md

@@ -0,0 +1,30 @@
+---
+type: task
+name: ask-table-needs-status
+---
+
+Read `context.current_db.table_purpose` to inform the suggestion.
+
+Agent determines a recommendation based on table type:
+- User/entity tables (users, admins, bots, products) → recommend YES
+- Event/log/audit tables (notifications, patterns, chart events) → recommend NO
+- Reference/lookup tables (country, strategy modules) → recommend NO (use is_deleted only if needed)
+- Join/pivot tables → recommend NO
+
+Ask the user exactly this question:
+
+"Should this table include a 'status' column? (0 = Active, 1 = Inactive)"
+
+Show recommendation: "Suggested: [YES/NO] — [reason based on table purpose]"
+
+Present options:
+1. Yes — add status INTEGER NOT NULL DEFAULT 0 CHECK (status IN (0,1))
+2. No — this table does not need a status column
+
+Wait for user selection.
+
+Store result in: `context.current_db.needs_status`
+- Option 1 → true
+- Option 2 → false
+
+Do not ask any other question in this task.

package/tasks/ask-table-purpose.task.md

@@ -0,0 +1,28 @@
+---
+type: task
+name: ask-table-purpose
+---
+
+Ask the user exactly this question:
+
+"Describe what this table will store and what it is used for."
+
+Example: "Stores registered users of the platform including their profile,
+contact details, and authentication information."
+
+Example: "Tracks trading bot configurations created by users, including
+their strategy settings and current running status."
+
+Wait for user input. Accept free text.
+
+Validate: must be at least 10 characters (not too vague).
+
+Store result in: `context.current_db.table_purpose`
+
+The agent uses this description to:
+- Suggest appropriate column names and types
+- Determine whether standard columns (status, is_deleted) are appropriate
+- Suggest relevant indexes
+- Write the comment header line in the SQL file
+
+Do not ask any other question in this task.

package/tasks/ask-table-seed-data.task.md

@@ -0,0 +1,44 @@
+---
+type: task
+name: ask-table-seed-data
+---
+
+Read `context.current_db.table_purpose` to determine suggestion.
+
+Agent evaluates:
+- Reference/master data tables (country, app_content, strategy_modules, admin) → suggest YES
+- User-generated data tables → suggest NO
+- Event/log tables → suggest NO
+
+Ask the user exactly this question:
+
+"Does this table need initial seed data?"
+
+Show recommendation: "Suggested: [YES/NO] — [reason based on table type]"
+
+Examples of seed data:
+- tbl_admin → default admin user record
+- tbl_app_content → About Us, Privacy Policy, Terms of Use entries
+- tbl_strategy_modules → strategy type entries
+
+Present options:
+1. Yes — this table needs initial data
+2. No — no seed data needed
+
+Wait for user selection.
+
+Store result in: `context.current_db.needs_seed`
+- Option 1 → true
+- Option 2 → false
+
+If true:
+  Ask: "Will this seed data be added directly in the table setup file,
+  or in a separate seed file in seeds/?"
+
+  Present options:
+  1. In the table setup file (for permanent reference/master data)
+  2. In a separate seeds/ file (for development/testing data)
+
+  Store: `context.current_db.seed_location` = "inline" | "seeds_file"
+
+Do not ask any other question in this task beyond seed location if applicable.

package/tasks/ask-target-service.task.md

@@ -0,0 +1,32 @@
+---
+type: task
+name: ask-target-service
+---
+
+Read `context.services` and list all initialized services.
+
+If no services exist:
+  Inform user: "No services have been initialized yet. Run @initialize-project first."
+  Abort task.
+
+If only one service exists:
+  Inform user: "Using service: [service_name]"
+  Store automatically: `context.current_action.service_name = <only_service>`
+  Return without asking.
+
+If multiple services exist:
+  Ask the user exactly this question:
+
+  "Which service would you like to target?"
+
+  List all services with their type and port:
+  Example:
+    1. auth (nodejs, port 1001)
+    2. ledger (nodejs, port 1002)
+    3. frontend (reactjs, port 3000)
+
+  Wait for user selection.
+
+Store result in: `context.current_action.service_name`
+
+Do not ask any other question in this task.

package/tasks/ask-test-type.task.md

@@ -0,0 +1,20 @@
+---
+type: task
+name: ask-test-type
+---
+
+Ask the user exactly this question:
+
+"What would you like to do?"
+
+Present options:
+1. Generate new test file for a module
+2. Run existing tests for a service
+
+Wait for user selection.
+
+Store result in: `context.current_test.type`
+- Option 1 → "generate"
+- Option 2 → "run"
+
+Do not ask any other question in this task.

package/tasks/ask-validation-library.task.md

@@ -0,0 +1,38 @@
+---
+type: task
+name: ask-validation-library
+---
+
+Ask the user exactly this question:
+
+"Which validation library would you like to use for this page?"
+
+Present options:
+1. Yup — schema-based validation, works well with React (recommended)
+2. Parsley — jQuery-based, good for legacy HTML forms
+3. React Hook Form — built-in validation with minimal re-renders
+4. Validator.js — lightweight utility functions, no form binding
+5. Custom — I'll write plain validation functions (no library)
+
+Wait for user selection.
+
+Store result in: `context.current_action.validation_library`
+- Option 1 → "yup"
+- Option 2 → "parsley"
+- Option 3 → "react-hook-form"
+- Option 4 → "validatorjs"
+- Option 5 → "custom"
+
+Show confirmation based on selection:
+- yup → "Yup will be added to package.json. Validation schema will be
+  built alongside the form."
+- parsley → "Parsley will be loaded via CDN or installed as a package.
+  Fields will receive data-parsley-* attributes."
+- react-hook-form → "React Hook Form will be added to package.json.
+  The form will use useForm() hook with register() on each field."
+- validatorjs → "Validator.js will be added to package.json. Each
+  field will be validated using validator utility functions."
+- custom → "No library will be installed. Validation logic will be
+  written as plain JavaScript functions in the component."
+
+Do not ask any other question in this task.

package/tasks/detect-repository-state.task.md

@@ -0,0 +1,92 @@
+---
+type: task
+name: detect-repository-state
+---
+
+## Purpose
+Inspects the repository to establish baseline state. Designed to be
+fast on repeat activations by checking context_version before doing
+any filesystem scanning.
+
+---
+
+## Step 0 — Check context directory exists
+
+Check if `.codeninja/context/` directory exists on disk.
+If it does not exist → create it now.
+This ensures write-context never fails due to a missing directory,
+even on the very first run before any context has been written.
+
+## Step 1 — Check context.json first
+
+Attempt to read `.codeninja/context/context.json`.
+
+If the file does not exist:
+  This is a fresh repository — must do full scan.
+  Set: `context.has_context_file = false`
+  Proceed to Step 2 (full scan).
+
+If the file exists:
+  Read `context_version` and `last_command_at` from it.
+  Set: `context.has_context_file = true`
+
+  Check `last_command_at`:
+  - If less than 60 seconds ago → the agent just ran a command.
+    The repository state has not changed externally.
+    Skip filesystem scan entirely.
+    Set all repository_state values from context directly:
+      `context.repository_state = "existing"`
+      `context.detected_services = [keys from context.services]`
+      `context.detected_db = (context.db.type != "")`
+    Return immediately. Do not scan filesystem.
+
+  - If more than 60 seconds ago → enough time has passed that the
+    developer may have made changes manually.
+    Proceed to Step 2 (full scan).
+
+---
+
+## Step 2 — Full Filesystem Scan
+
+Only runs on first activation or when time threshold exceeded.
+
+Check for:
+- Any folders with `package.json` or `app.js`
+- Any `database/` folder
+- Any `src/main.jsx` or `src/index.jsx` (React apps)
+
+Set context values:
+- `context.repository_state` = "empty" | "existing"
+- `context.detected_services` = [] (list of folder names that look
+  like services — folders containing app.js or package.json)
+- `context.detected_db` = true | false (database/ folder exists)
+
+---
+
+## Step 3 — External Modification Check
+
+If `has_context_file` is true AND full scan ran:
+  Compare `context.detected_services` from scan against
+  `Object.keys(context.services)` from context.json.
+
+  If they differ:
+    Inform user:
+    "⚠ Repository has changed since last agent run.
+    Services on disk: [detected_services list]
+    Services in context: [context.services keys]
+    Consider running @sync to bring context up to date."
+
+---
+
+## Step 4 — Missing Context Warning
+
+If `has_context_file` is false AND `repository_state` is "existing":
+  Inform user:
+  "Repository has files but no context.json found.
+  Run @sync to build context from existing code."
+
+---
+
+## Return
+
+Return: `repository_state`