Tally Robot Script

The all-in-one Visual Studio Code extension for writing, reviewing, navigating, and executing Tally Robot Scripts.

TallyRobot Script AI is a comprehensive Visual Studio Code extension designed specifically for Tally Robot Script development and automated testing workflows. It combines intelligent code assistance, script validation, map file navigation, execution capabilities, and AI-powered assistance into a single integrated experience.

Whether you are creating new automation scripts, reviewing existing ones, searching map libraries, or executing tests through Tally Robo 2.0, TallyRobot Script AI helps improve productivity, reduce errors, and streamline development workflows.

Features

• TallyRobot Script AI command box with dual modes: - Local Mode — Review, Search Maps, Open Map, Run Scripts

• TallyRobot AI Mode — AI-powered assistance with email-OTP sign-in

• llamaserver — LlamaServer (Self Hosted AI)

Chat with Qwen, an open-source LLM running privately on Tally Solutions' OCI infrastructure.

No sign-in or Cognito session required No token limits Your data never leaves the network If your response contains code, it automatically opens as a new file in the VS Code editor with the correct language (Python, TypeScript, SQL, etc.) Status indicator shows whether the server is reachable

• Scriptwriter Auto mode — Work in Tally, get the Robot Script written for you live from the execution log • Script Discovery — Search across all indexed scripts using plain English, powered by semantic vector search via AWS S3 Vectors • Script review with inline fix suggestions • Multi-script review — Review all scripts in a folder at once

• Search in map files • Open map file • Ctrl+Click navigation — Jump to map files and loop parameters directly from script lines • Context aware autocomplete • Code Snippets and hover information • Help comment insertion — Add structured help comment headers to scripts • Slash commands — /review, /discovery, /mapsearch, /openmap, /run, /help, and more • Template scripts • Syntax highlighting for Tally Robot scripts • Formatting for Tally Robot scripts • Syntax highlighting for map files • Error detection • Enhanced UI/UX — Sidebar panel theme auto-syncs with your VS Code colour theme • Run scripts directly via Visual Studio Code (powered by Tally Robo 2.0) • Mock API Log Extraction Tool

Release 5.0.0 Updates:

Mentor Mode

Mentor Mode turns the command box into a live scripting coach that follows your cursor as you write a TallyRobot script — explaining what each line does, suggesting the likely next step, and quietly reviewing your work in the background.

Everything is computed locally against the extension's .map/pattern indexes in milliseconds — no AI Mode login, no network round-trip.

How to Use

1. Open a TallyRobot script (.txt) file, then switch the provider chip in the command box to Mentor.
2. Starting a new script? Mentor leads with a header card — click Insert header to drop in a fully annotated header skeleton, each field explained with its own ;; comment.
3. As you type or move the cursor, Mentor keeps up to four suggestion cards in sync — matches for a ;; comment intent you just wrote, or the likely next line based on how other scripts continue from here. Click Insert to drop a suggestion in at the cursor; it's removed from the panel automatically once inserted.
4. Pause for about two minutes and Mentor nudges you with a background standards review — errors, warnings, and how many it can auto-fix. Click the card to open the full Script Review panel for the details.

Tips

• Suggestions show their provenance — e.g. "Seen in 42 scripts" — so you can judge how well-trodden a suggested line is before inserting it.
• Inserted multi-line blocks arrive pre-commented — Mentor writes a plain-English ;; explanation above each line for you.
• The idle review reuses the same engine as the standalone Script Review feature, so the counts and fixes you see always match what a full review would report.
• No robot script open yet? Mentor still greets you and explains itself — switch providers any time to bring it back.

Scriptwriter

Accessed via the /Scriptwriter command in the command box. It has two modes.

Manual mode (No changes since 4.2.0)

Script writer — Launches Tally with a custom TDL and monitors execution logs in scriptwriter.log in real time.

• When we left-click on a field in Tally, the field name will be copied to the clipboard.
• Users can click Map Search in the command box to search within the map file lines.

Auto mode — log to script converter (*New)

Launches an Execflow build of Tally and writes the Robot Script for you, live, while you work. You do the steps in Tally normally; the script appears in your target file as you go. There is no Generate button.

How it works

1. Pick your Tally Execflow executable and choose where the script should go.
2. Tally starts with the /execflow flag and writes its own trexecuflow<PID>.log.
3. The extension verifies the build is genuinely Execflow, then re-reads that log every 1.5 seconds and converts it into a Robot Script.

Generate into

Option	What it does
Append	Inserts the generated block into the file open in the editor, at your cursor line. Existing content stays. Needs a file open that sits beside the release lib.maps.
Create New	Creates a new file in a folder you pick.

The target folder must sit next to lib.maps — for example Release_3.0\Company. The release root (Release_3.0) and the maps library itself (Release_3.0\lib.maps) are both rejected, with a message explaining why.

What you get

• A filled-in header block — company name and number, release used, created on/by — read from the log.
• For an existing company, a MapFile | SelectCompany | | line. If the company needs a login, the username and password are carried onto it: MapFile | SelectCompany | | username=a # Password=a. TallyVault passwords work the same way.
• Startup noise removed — company selection, migration and login screens at the start, plus arrow keys, Ctrl+Shift+*, Alt+Shift+*, Alt+0, and Enter presses on fields you did not edit.
• Menu paths and multi-row macros collapsed into single lines where a map row covers the whole sequence.

Nothing is silently dropped. If a step has no matching row in lib.maps, a mapline is invented from what actually happened, saved to temp.map beside your maps, and flagged in the script with a ;; REVIEW: comment so you can check it.

Map file priority — assets/mapPriority.md lets you say which map file to prefer on a given screen when several would match. The order written is the priority. The file refreshes itself from S3 in the background.

Test Deployment (Jenkins)

Test Deployment lets you trigger and track a Jenkins BaseDeployment build — generate, verify, or deploy a Base — directly from the command box, with parameters auto-filled from whichever script you have open.

Sign in once with your Jenkins username and API token; the extension talks to Jenkins through its MCP server and polls the triggered build until it finishes, streaming live pipeline progress back into the panel.

How to Use

1. Type /testdeployment in the command box — or a natural-language request like "generate the base" or "deploy TP_9.2" — to open the Jenkins Deployment panel.
2. Sign in with your Jenkins username and API token (stored securely for future sessions).
3. With a script open, its execution name, module, and sub-repo are auto-filled into the parameter form; release, build folder, and alternate tags are derived automatically where possible.
4. Review the essential parameters (e.g. Result Tag) — everything else sits collapsed under Advanced. Click Trigger Build.
5. Watch live progress in the panel — a stage-by-stage pipeline view updates automatically until the build finishes. Click any stage to view its logs inline.
6. On a successful generation/verification, links to the Base Generation report and Test Report appear directly on the result card.

Tips

• Natural language works for the details too — e.g. "verify base for TP7.1PB using build 12345" picks out the build number, release, and result tag automatically.
• Before triggering, the extension checks that your script has actually been promoted to the OCI bucket first — you'll be told if it hasn't, instead of wasting a build.
• Ask "list jobs" or "show me all jobs" to see what's available.
• Ask "view the base" to pull up the existing Base Viewer result before generating a new one.
• Sign out from the panel any time to switch Jenkins accounts.

Release 4.2.0 Updates:

llamaserver — LlamaServer (Self Hosted AI)

Chat with Qwen, an open-source LLM running privately on Tally Solutions' OCI infrastructure.

No sign-in or Cognito session required No token limits Your data never leaves the network If your response contains code, it automatically opens as a new file in the VS Code editor with the correct language (Python, TypeScript, SQL, etc.) Status indicator shows whether the server is reachable

UI Changes:

1. Mode Switching (Discovery/AI Features Unified with Local)

• Slash commands (/discovery, /review, /mapsearch, etc.) now show in both Local and AI mode — previously AI-only or Local-only commands were hidden depending on the active mode.
• Selecting an AI-only command (e.g. /discovery) while in Local mode now auto-switches the session to AI mode, walking you through login first if not signed in, then resuming straight into the feature — no more "switch modes manually first" dead end.

2. "What's New" Notification Panel

• New card-based panel shown only when a session has zero messages, with content driven by whatsnew.md (section-per-card format with display, tag, icon, and try metadata) — announcements can be added or toggled without a code change.
• Multiple cards auto-rotate every 3s, pause on hover, with clickable dot indicators and hover-reveal prev/next arrows for manual navigation.
• Each card can carry a Try now button that runs the exact command as if typed by the user (e.g. jumps straight into Discovery).
• Content is fetched from a public OCI Object Storage bucket (TallyRobotScriptExtension/discovery/whatsnew.md) with conditional-GET/ETag caching, cached in globalStorage, falling back to a bundled seed copy if offline or before the first successful sync.
• The panel is centered in the chat area rather than pinned to the top.

3. Extension Update Banner

• Compares the installed version against the public VS Code Marketplace and shows a banner (gradient, pulsing badge icon) whenever they differ — positioned just above the message composer.
• One-click Update now installs the latest version via workbench.extensions.installExtension, then prompts to reload the window to activate it.
• Dismissing the banner is session-only — it reappears on the next chat/session open rather than being permanently silenced.

TallyRobot Script AI Command Box

The extension includes a dedicated TallyRobot Script AI sidebar panel accessible from the activity bar. The command box lets you interact with the extension using natural language commands from a single place — no need to hunt through menus or the command palette.

You can type commands such as:

Command	What it does
review	Reviews the active script and shows issues with fix suggestions
search maps <query>	Opens the map search panel pre-filled with your query
open map <name>	Opens the specified map file from the current maps library
settings	Opens the extension settings
maps path	Updates the maps library directory path
repo path	Updates the code.robotscript repository path

The command box also maintains a chat history across sessions. Each session stores your commands and bot responses, and you can switch between sessions or start a new one using the session header controls.

Tips:

• Type help to see a quick summary of available commands.
• Review artifacts (completed reviews) are saved in the session history as clickable cards — click any card to re-open that review.
• Use the ⊕ button in the session header to start a fresh session, or the ☰ button to browse and switch between past sessions.

TallyRobot AI Mode

TallyRobot AI Mode provides AI-powered assistance with secure email OTP authentication and AWS Bedrock integration.

Features:

• Dual Mode Support • Local Mode — Review, Search Maps, Open Map, Run Scripts • TallyRobot AI Mode — AI-powered assistance

• Email OTP Authentication • Passwordless sign-in • One-time password delivered to work email • 30-minute session • Automatic sign-out on expiry • Session persists across VS Code restarts

• Provider Toggle • Switch between Local and TallyRobot AI modes • Visual status indicators • Session countdown timer

• Seamless Integration • Local commands continue to work in both modes • Session history preserved • Automatic bearer token handling

• Getting Started

1. Run /login
2. Enter your work email
3. Enter the OTP received
4. Start using AI-powered assistance

Features:

• Dual Mode Support:

  • Local Mode — All existing features (Review, Search Maps, Open Map, Run Scripts)
  • TallyRobot AI Mode — AI-powered assistance with secure authentication

• Cognito OTP Authentication:

  • Email-based OTP login (no password required)
  • 30-minute session with automatic expiration
  • Live countdown timer in the auth bar
  • Automatic token refresh before expiration

• Seamless Integration:

  • All Local commands work in both modes
  • Chat-based OTP entry
  • Session history preserved across mode switches

Security:

• All credentials are stored securely in the VS Code settings
• Bearer tokens are automatically managed and refreshed
• OTP sessions expire after 30 minutes
• Email domain restriction available for enterprise deployments

Troubleshooting:

If you encounter issues with TallyRobot AI Mode:

1. Restart your VS Code app and press Ctrl + Shift + P -> Restart development host
2. Contact TestOps for any other queries

Script Discovery

Script Discovery lets you find the right TallyRobot script using plain English — no need to remember file names or browse folders. Describe what you're looking for and Discovery returns the most relevant scripts ranked by semantic similarity.

Your query is sent to a secure server that turns it into a vector and matches it against the script index stored in AWS S3 Vectors. Cognito authentication (the same login used for TallyRobot AI Mode) gates all Discovery queries.

How to Use

1. Switch to TallyRobot AI Mode (Cognito login required)
2. Click "Click here to open Discovery Chat" in the panel, or type /discovery in the command box
3. Enter a plain-English query — for example: "performance test for ledger" or "voucher entry for sales"
4. Results appear as cards showing the script name, module, release, and a relevant snippet
5. Click any card to open the script file directly in the editor — the matching lines are highlighted automatically
6. Click Show More to load additional results

Tips

• Use natural language — queries like "Navigate to balance sheet" or "perf script to Navigate to sales report"
• If your query mentions performance, Discovery automatically filters to performance scripts only
• Click Show More to load additional results beyond the initial set (if any)
• Your previous Discovery searches are saved in the session — switching sessions restores your earlier results

What's New in Discovery

• Mention a release or module in your query — ask things like "script from Rel 7.1" and Discovery understands the release/module and narrows the search before looking up results.
• Fresh results on repeat — re-run the same query and Discovery returns different scripts instead of the same list.
• Scripts are tagged with their suite — if a script belongs to a whitelisted suite, that suite is attached to the script during indexing.
• Commented-out code is ignored — commented code lines in scripts are skipped during indexing.
• Jira ID tagging — if a script's comments mention a Jira ID, it's attached to the script so results carry that reference.
• Smarter de-duplication — duplicate comments are now considered per Release + Module instead of globally.

TallyDataOps

TallyDataOps lets you run bulk Migrate, Reindex, Repair, and Load Company operations across many Tally companies in one go — driven entirely from a single Excel job list, instead of repeating the same steps manually in Tally for each company.

Each row in the Excel sheet describes one company/job (source data path, company name, credentials, and which operations to run), and the extension automates Tally in the background to process every row, copying the resulting data to your specified output folder.

How to Use

1. Type /TallyDataOps in the command box (natural language like migrate, reindex, or repair also opens the panel), or launch it from the Help panel, or via the Command Palette.
2. Click Browse to select your job Excel file, or click Download Sample Excel to get a template showing the required column layout.
3. Choose the number of Concurrent Jobs (1–5) to control how many companies are processed in parallel.
4. Click Run to start processing. Progress is shown via a progress bar (N / Total rows completed) and a live, color-coded Execution Log (copy, launch, success, warning, and error events).
5. Click Stop at any time to cancel — in-flight Tally processes are terminated and partial data is cleaned up automatically.

Tips

• The first sheet's cell B1 must contain the full path to your tally.exe / tallyprime.exe; data rows start from row 3.
• Each row can independently request Reindex and/or Repair alongside the Migrate step — set the corresponding flag columns.
• Specify an output folder per row to automatically receive the processed company data once a job completes.
• After the run finishes, a consolidated summary (overallmig.txt) merges the results of every row for quick review.
• Use the Clear link in the Execution Log to reset the log view between runs.

Script Review

The Review feature analyses your active Tally Robot Script and surfaces errors, warnings, and informational hints — all within the sidebar panel. Each issue is shown with a description and, where applicable, a one-click Accept Fix button to apply the suggested correction directly in the editor.

To trigger a review:

• Type review in the command box, or
• Click the Review Script button (checklist icon) in the editor title bar, or
• Right-click in the editor and choose Review Script from the context menu.

Tips:

• Use Accept All Fixes to apply every auto-fixable suggestion in one go.
• Use Undo Fix to revert the last applied fix.
• Click Go to Line on any issue to jump directly to the relevant line in the editor.
• After a review is complete, a summary card (showing error / warning / info counts) is saved in the session history. Click it at any time to re-open that review result.

Multi-Script Review

Need to review an entire folder of scripts at once? Multi-Script Review analyses all scripts in a selected folder and generates a consolidated review report.

How to Use:

• Click on the "+" icon at the bottom left of the chat panel • Attach the folder in which you want your scripts to be reviewed • Type review or /review -> Enter • Alternatively you can select multiple files and review them all at once upon clicking

Tips- Type @ in the prompt to mention files or folders inline. Picker appears as you type. You can attach multiple folders and multiple scripts for review

What You Get • Consolidated report • Per-script analysis • Clickable issue cards • Summary statistics

Ctrl+Click Navigation (Go To Definition)

Navigate directly to map files and loop parameters.

• Ctrl+Click Target Action
• Map File Open map file
• Loop Parameter Jump to map entry
• File Path Open referenced file

Tips • Hover for preview • Works with all configured map libraries

Search In Map Files

You can conveniently search for key sequences, report names, or keywords within the map files of the active maps library by selecting the "Search In Map Files" option from the context menu (right-click in the editor). Additionally, the "Search In Map Files" command is accessible through the command palette (Ctrl+Shift+P), or by typing search maps <query> in the command box.

The map search panel opens inside the sidebar and provides:

• Release selector — switch between available releases or browse a custom maps folder without leaving the panel.
• Search scope — narrow results to All Fields, Map Names only, or Loop Parms only.
• Mapfiles filter — expand the Mapfiles panel to select or deselect individual .map files to search within.

Tips:

• To search for a keyword within the editor, highlight it with the cursor, right-click to access the context menu, and choose the "Search In Map Files" option to locate that keyword in the map files.

• You can press Enter (or click a result card) to insert the selected map entry into your script in the correct MapFile | <file> | <mapname> | format.

• The file selection is remembered across sessions — the same set of map files will be pre-selected the next time you open the search panel.

• Use the Show Selected button in the Mapfiles panel to quickly review which files are currently included in the search.

Open Map File

Access any map file from the current maps library swiftly by selecting the "Open Map File" option in the context menu (right-click within the editor). You can also utilize the "Open Map File" command from the command palette (Ctrl+Shift+P).

Tips:

• To open a map file name that is currently displayed in the editor, you can easily highlight it with your cursor, right-click to access the context menu, and choose the "Open Map File" option for a quick access to that map file.

Context Aware Autocomplete

Receive smart autocomplete recommendations while you type, tailored to the context of various parameters such as the active map file library, the map file in use, and the current map name.

Code Snippets and Hover Information

• Utilize code snippets to enhance speed and efficiency while scripting.

• Hover over script components to gain quick insights into their functionality.

Template Scripts

If your team often utilizes a specific script pattern or style, you can now designate that script as a template by commiting it in the code.robotscript/templates/ directory. After this is completed, team members will be able to use the "Insert Template Script" autocomplete feature in the editor, allowing for quick insertion of the template script and enhancing efficiency.

Formatting

This extension includes a custom formatter specifically designed for Tally Robot Scripts, allowing you to enhance and beautify your scripts with a single click. To format your script, right-click within the script area to access the context menu and choose the format document option. Upon your initial formatting attempt, you may encounter a popup indicating that no default formatter has been established for Tally Robot Scripts. If this occurs, just click on the "Configure" or "Select Formatter" option in the popup and choose the tallyrobotscript formatter from the dropdown menu.

Compilation and Error Detection

Compile your scripts to detect all errors at once, thus avoiding the tedious process of running the script, facing errors, correcting them, and re-executing it to pinpoint individual issues. Furthermore, obtain valuable suggestions for possible solutions to tackle each identified problem. To compile your script, right-click within the script and choose the "Compile Robot Script" option from the context menu. You can also utilize the "Compile Robot Script" button located on the top bar or editor title bar.

Script Execution

Run your scripts effortlessly with a single click in one of the two available modes, powered by Tally Robo 2.0. The base generation mode allows for the rapid creation of a local base complete with test case screenshots, while the custom config mode launches the Tally Robo 2.0 GUI, enabling you to tailor all execution parameters to meet your specific needs prior to script execution.

You can initiate your script using either the buttons on the top bar/editor title bar or by selecting the execute commands from the context menu (right-click) within the active editor.

FAQ (Frequently Asked Questions)

1. How to know/check which maps library is currently in use for providing the autocomplete suggestions?

The most straightforward method to identify the active map file library is to check the maps library indicator located on the right side of the status bar (the bottom bar in Visual Studio Code). For instance, if the indicator displays Release_6.0, it indicates that the map files from code.robotscript/Release_6.0/lib.maps (on your local machine) are currently utilized for autocomplete functionality.

When dealing with a script file from the code.robotscript repository, the extension automatically adjusts the maps library based on the context. For instance, if you access the script located at code.robotscript/Release_6.0/User_Sagar/testscript.txt, the extension recognizes it as a script within Release_6.0 and subsequently changes the maps library to code.robotscript/Release_6.0/lib.maps.

If a script is located outside the code.robotscript repository, the extension will continue to utilize the currently employed maps library unless the user explicitly modifies this setting.

In cases where the automatic switching of maps libraries does not function correctly, defaults to an incorrect library, or if the user prefers to manually select a maps library, the user has the option to utilize the "Change Maps Library Directory Path" feature found in the context menu (right-click) of the editor. Additionally, the user can simply click on the maps library status bar indicator and choose the preferred maps library from the dropdown menu.

2. Automatic maps library switching is not working properly for me. What to do?

This issue may arise from the local code.robotscript repository path not being configured correctly in the extension settings. Initially, when you run the extension for the first time, you will receive a prompt to specify the code.robotscript path on your system. However, if you inadvertently closed that prompt or if the path has changed since then, you can manually update it by selecting the "Change code.robotscript Path" option in the editor's context menu (right-click). Alternatively, you can also access this option through the Visual Studio Code command palette (Ctrl+Shift+P).

3. I have opened a Tally Robot Script(.txt) file but I am not at all getting any syntax highlighting or autocomplete features. What could be the issue?

Visual Studio Code may be identifying the file with a .txt extension as a standard text file rather than a Tally Robot Script file. To resolve this issue, you can manually adjust the file's language setting from the status bar at the bottom of Visual Studio Code to Tally Robot Script. Another option is to utilize the "Change Language Mode" command available in the command palette (Ctrl+Shift+P) to modify the file's language.

4. I get errors when trying to execute my script via Visual Studio Code. What should I do?

Ensure that the Tally Robo 2.0 path is correctly configured in the extension settings. Initially, when you attempt to run a script for the first time, a prompt will appear asking you to specify the Tally Robo 2.0 executable path on your system. If you inadvertently closed that prompt or if the executable path has changed, you can manually update the path by selecting the "Change Tally Robo 2.0 Executable Path" option from the editor's context menu (right-click). Alternatively, this can also be accomplished through the Visual Studio Code command palette (Ctrl+Shift+P).

5. I find myself using certain commands often and want to add keyboard shortcuts for them. Is it possible?

Visual Studio Code provides users with the ability to create custom keybindings for nearly any command or action within the editor. To access the keyboard shortcuts interface, you can open the command palette by pressing Ctrl+Shift+P and selecting "Preferences: Open Keyboard Shortcuts." In this section, you can search for the desired command and customize its shortcut according to your preferences.

6. The command box says "Command not recognized." What commands can I use?

The command box understands natural language variations of the following actions: review, search maps <query>, open map <name>, run base, run gui, settings, maps path, repo path, and tally robo path. You can also type help to get a quick reminder of available commands directly in the chat.

7. Can I search map files for a specific release without changing the global maps library?

Yes. The map search panel has its own release selector in the header. You can switch to any available release (or browse a custom folder) directly within the search panel — this only affects the search scope and does not change the global maps library used for autocomplete.

Feature Summary

Feature	Access
Command Box	Sidebar Panel
Script Review	Command Box · Context Menu · Editor Toolbar
Multi-Script Review	Command Box · Folder Context Menu
Ctrl+Click Navigation	Ctrl+Click / F12
Map Search	Command Box · Context Menu
Open Map File	Command Box · Context Menu
Help Comment	Context Menu · Command Palette
Autocomplete	Automatic
Formatting	Context Menu
AI Mode	login
OTP Authentication	In-chat OTP
Theme Synchronization	Automatic
Mock API Extraction	Command Palette
Script Discovery	/discovery command · "Click here to open Discovery Chat"
Scriptwriter (Manual / Auto)	/scriptwriter command
Script writer	/Scriptwriter command