This page explains what each family solves, when it is useful, and how it fits into the live SindByte runtime.
Session flow, thinking extraction, voice features, LMChat handoff, and the multi-model operating pattern.
Model selection, IQ helpers, transcription, attachments, and the lighter one-model execution surface after Dialog-LAB.
12 venues, tool stack, DecisionMenu flow, credential checks, and a paper-first start pattern.
File, data, system, memory, clipboard, and desktop automation families work as the practical Windows bridge.
These are the first surfaces to validate when you want offline utility.
Cloud, Gmail, M365, Teams, and parts of trading remain dormant until credentials and registration settings allow them.
Absence in a client does not automatically mean the feature is missing from the product.
Dialog-LAB, LMChat, Timer screens, Config UI, and Trading Hub are product forms around the MCP server, not just bare tool names.
The website now documents both tool families and operator-facing forms.
Connect IQ runs and cloud-backed routes to the right provider and model stack.
Turns one good prompt into a scheduled operating routine with stored follow-up behavior.
Shows that the reasoning families also have a live monitoring surface, not just callable names.
Manual trading and paper-first execution sit beside the audited MCP catalog as real product forms.
Generated with the live SindByte server to make the audited runtime surface easier to scan at a glance.
A second generated visual that shows how per-tool workflows often move from file intake to structure, validation, scheduling, and visual output.
Use this grid when you already know the family and need the right tool fast. In short registration, the same target names are called through ToolCatalog.CallTool.
ToolCatalog.CallTool
Load local text or source files into an agent flow.
Best when you need a file body before summarizing, patching, or extracting data.
Pull one path or sub-tree out of a larger JSON document.
Best when API output is too large and only a few keys matter to the next step.
Capture a live desktop, window, or workflow state for proof and troubleshooting.
Best when documentation needs the actual UI, not a paraphrase.
Create an explanatory visual, hero asset, or diagram from a prompt.
Best when one picture communicates a feature faster than text alone.
Adapt screenshots and generated images to website-friendly sizes.
Best when the original capture is correct but too large for a page layout.
Run an answer through a structured validation pass before publishing it.
Best when you want a second pass for factual, logical, or code quality checks.
Store a prompt-driven schedule for once, interval, daily, or weekly runs.
Best when a good prompt should become an operating routine instead of a manual repeat task.
Execute Windows-side automation or inspection tasks from the MCP layer.
Best when the workflow needs controlled system introspection beyond file access alone.
Read, write, move, archive, and inspect files. Useful for code work, report generation, and controlled automation on the local machine.
Client fit: LM Studio, Roo Code, VS Code add-ons, Codex.
Turn text, CSV, and JSON into a usable working set. Good for pre-processing before reasoning or output generation.
Client fit: all MCP clients with structured output.
Store and recall intermediate state across multiple steps. Strong in agent loops, review pipelines, and long-running assistant sessions.
Client fit: especially helpful in multi-step workflows.
Formula resolution, optimization, integrations, simulations, and sequence helpers for engineering, forecasts, and validation.
Client fit: pairs well with IQTools.
Multi-step reasoning with MultiVote, Validate, Debate, and SelfCritique to raise answer quality beyond single-pass completion.
Client fit: LM Studio, OpenAI-backed flows, Dialog-LAB.
Schedule prompts, recurring checks, and repeatable reminder logic. Helpful for routine evaluations and monitoring tasks.
Client fit: directly via server GUI and MCP calls.
Desktop automation, screenshots, window targeting, and image workflows. This is where Windows interaction becomes visible.
Client fit: strong for practical automation in all MCP hosts.
Process, service, shell, registry, and environment operations with security levels to keep powerful actions under control.
Client fit: admin, ops, and local integration flows.
Fetch and prepare web material for research flows, follow-up extraction, and reasoning pipelines.
Client fit: research, summarization, controlled fetch chains.
Read and write the clipboard for fast transfer between AI and applications without file management overhead.
Client fit: copy/paste heavy workflows.
Plan and manage tasks with priority and status where the active build exposes this family.
Client fit: project and workflow management in SLL scenarios.
Patch and diff oriented text changes for review, cleanup, and precise edit flows.
Client fit: content creation, code review, redline work.
Multi-model conversations with thinking extraction, voice input/output, and session logging.
Client fit: LM Studio with multiple active models.
12 trading venues, API-key verification, paper/live modes, and IQ-assisted session flow.
Client fit: trading research and execution support.
The lists below show documented tool names. Credential-bound and build-specific families can remain invisible at runtime until profile and registration match the deployment.
ReadFile, WriteFile, AppendToFile, Delete, Move, Copy, Zip, Unzip, GetDirectoryListing, List, GetInfo, GetLines, FindText, CreateDirectory, SearchInFiles, FileExists
CsvToJson, JsonToCsv, JsonSelect, TableToJson, FindText, Replace, RegexSearch, RegexReplace, Split, Join, GetLines, SortList, ParseKeyValue, GetInfo, LoopStart, LoopStep, Chunk, Echo, NormalizeNewlines, NormalizeWhitespace
Store, Recall, Delete, ListKeys, Format, FromFile, ToFile, FromClipboard, ToClipboard, GetLines, JSON_Escape, JSON_Unescape
ResolveFormula, LoopStart, LoopStep, LoopCalc, MonteCarlo, Optimizer, RootFinder, Integrate, SmartSequence, BigIntCalc, BinaryCalc
ChainOfThought, MultiVote, SelfCritique, CrossResolve, Debate, Ensemble, Validate, BetterPrompt, GetResult, SelectModel, GetModel, TreeOfThought, MetaReason, Socratic, ExpertPanel, RedTeam, Hybrid, Iterate
SetTimer, GetTimer, ListTimers, FindTimer, UpdateTimer, ClearTimer, TestPrompt
Screenshot, MouseClick, SimKeyPress, ListWindows, GetWindowState, SetWindowState, LocateAppWindow, LocateChildWindow, LocateGUIElement, SetPosAndSize, PictureResize, PictureConvert, PictureCrop, GenerateImage, PictureEdit, AnalyzeScreenshot, LMStudioText
ListProcesses, ListServices, GetServiceDetails, StartService, StopService, RestartService, ExecuteCommand, RunPowerShell, RunBatchFile, GetCommandResult, ConfirmOperation, ListRegistrySubKeys, ListRegistryValues, CreateRegistryKey, DeleteRegistryKey, WriteRegistryValue, DeleteRegistryValue, GetEnvironmentVariable, GetInstallPath, IdentifyConhostParent, IdentifySvchostService
FetchText, FetchHeaders, FetchMany, ExtractText, Search
Read, Write, Clear, GetFormats
CreateTask, UpdateTask, ListTasks, GetTask, DeleteTask
ApplyDiff, ApplyUnifiedDiff
Multi-Model Chat, Thinking Extraction, Text-to-Speech, Voice Input, Conversation Logging, Round Management
Kraken, Binance, Coinbase, OKX, KuCoin, Bitget, ByBit, Bitpanda, Alpaca, IG, Interactive Brokers, Polygon
apply_diff_edit, spr_server_assistant
Excel, Word, PowerPoint, Outlook, Publisher automation
GenerateImage, PictureEdit, TextToSpeech, Kraken, Binance
ListMessages, Search, GetMessage, SendMessage, Labels
Microsoft 365 mailbox, calendar, contacts integration
Microsoft Teams chats, channels, meetings, contacts
MultiVote, Consensus, Scanner, Learning
DecisionMenu - AI Trading Decision Workflow