From data analysis and report writing to database queries, server management, email integration, and workflow execution — it handles complex tasks for you.
Reviews safety before execution and blocks hallucinations and dangerous operations.
Records every step and result after execution.
Handles complex tasks using multiple tools including files, databases, APIs, and server commands.
Before execution, DACS reviews the plan and HazardGuard blocks hallucinations and dangerous commands. Only reviewed tasks are executed, and every step and result is recorded.
wiiiv
✓
Request Analysis
Sales analysis + report email
✓
Execution Plan
DB query → analysis → report → email
✓
DACS Review
Architect ✓ Reviewer ✓ Adversary ✓
✓
HazardGuard
No risks detected — PASS
✓
Execution Complete
4 tasks completed, audit log saved
Sales analysis report created and emailed to the team lead.
📋 Audit log saved
Data Analysis
Analyzes raw data and delivers verifiable conclusions
Import raw data from CSV, spreadsheets, and databases, clean and combine it, analyze with code, and complete visualizations and reports — all within the conversation.
👤
Collect basic statistics on global carbon emission trends. Save as raw.md.
Confirmed raw.csv file. Contains 34 years of global CO₂ emission data from 1990 to 2023. Compiling basic statistics.
importpandasaspddf=pd.read_csv('raw.csv')world=df[df['Entity'].eq('World')][['Year','Annual CO2 emissions']]world=world.sort_values('Year').reset_index(drop=True)
MD Global Carbon Emission Trends Basic Statistics raw.md
👤
Check if the data is sufficient, and if not, search for and supplement additional data.
After reviewing raw.md, the data is sufficient for basic trend analysis. However, for enrichment, it would be good to include year-over-year changes and the last 5 years together.
Global Carbon Emission Trends (1990-2023) — Mt CO₂
IMG world_co2_trend_line.png
Additionally calculated decade-over-decade change rates. The rate of increase has been slowing since the 2010s.
Decade-over-Decade Change — 4 rows × 3 cols
Period
Change (Mt)
Rate
1990→2000
+3,651.2
+16.1%
2000→2010
+7,456.8
+28.3%
2010→2020
+1,461.5
+4.3%
2020→2023
+2,935.8
+8.3%
Trend Analysis — Surged +28.3% in the 2000s due to China's industrialization. Slowed to +4.3% in the 2010s with renewable energy expansion and developed-nation reductions. After a pandemic dip (-5.2%) in 2020, emissions rebounded to a record high in 2023.
👤
Compare per capita emissions too. We need to see responsibility relative to population.
Saved a bar chart comparing carbon emissions by major countries. Here is the data for the top 6 countries as of 2023.
Major Countries 2023 Emissions — 7 rows × 3 cols
Country
Year
CO₂ (Mt)
China
2023
12,172.0
US
2023
4,918.4
India
2023
3,062.8
EU(27)
2023
2,485.0
Russia
2023
1,733.1
Japan
2023
986.9
Major Countries CO₂ Emissions (2023, Mt)
12,172China
4,918US
3,063India
2,485EU(27)
1,733Russia
987Japan
IMG major_countries_co2_bar.png
Combined population data to calculate per capita CO₂ emissions. The ranking differs significantly from total emissions.
Per Capita Analysis — The US (14.5t) is 3x the global average (4.7t). China, #1 in total emissions, ranks 3rd per capita at 8.5t. India is 3rd in total but only 2.1t per capita, below the global average. Viewed through per capita metrics rather than totals, the historical responsibility of developed nations becomes clearer.
👤
Analyze the main causes and environmental impacts of carbon emission increases.
Here is your requested analysis of main causes and environmental impacts of carbon emission increases.
Main Causes — Growing energy demand (industrialization, urbanization, rising living standards), fossil fuel dependence (coal, oil, gas-centered energy supply), industrial production expansion (steel, cement, chemicals), transportation growth (road, aviation, maritime), urban sprawl and rising consumption
Environmental Impacts — Rising global average temperature (increased radiative forcing), intensifying extreme weather (heatwaves, droughts, heavy rainfall, typhoons), glacier and sea ice loss with sea level rise, ocean acidification (CO₂ absorption lowering pH), ecosystem disruption (habitat migration, biodiversity loss)
👤
Summarize future outlook and major countries' carbon neutrality policies.
Sector Analysis — Power and heat generation is the largest source at 40% of total emissions. This is especially pronounced in China and India where coal-fired power makes up a large share. The transport sector (16.2%) has yet to transition to a declining trend despite EV adoption.
👤
Create a docx report with all the analysis above. Include charts, insert tables, and cite sources.
DOCX report generated. Contains 2 charts, 9 tables, and source citations in a 16-page document.
Verification Results — 6 rows × 2 cols
Item
Result
Table count
9
Primary sections
8
World trend chart
Included
Country comparison chart
Included
Source citations
Included
Image relations
2
DOCX carbon_emissions_analysis_report.docx
Data Inspection & Enrichment
Checks for missing values, ranges, units, and data sufficiency. Supplements with additional collection or external data combination when needed.
Code-Driven Analysis
Writes and executes analysis code in Python, R, and SQL. Covers cleaning, statistical testing, regression, time series, and correlation analysis.
Verifiable Computation
Preserves the data used, formulas, execution code, and intermediate results. Anyone can reproduce the same process.
Visualization & Reports
Generates appropriate charts — bar, line, scatter — and composes interpretations and sources into finished report documents.
Result Re-examination
Re-checks through DACS consensus whether the requested analysis and deliverables are accurately reflected. Verifies numerical errors and interpretation bias.
DACS
Before AI executes, the independent oversight system DACS verifies it
DACS is an independent oversight mechanism separated from the AI executor. Three AI personas — Architect, Reviewer, and Adversary — examine the feasibility, accuracy, and safety of the user's execution plan from different perspectives.
If even one objects, the plan is not approved. The plan is revised based on rejection reasons and reviewed again; only plans with unanimous agreement proceed to execution.
DACS Consensus
"Please query the employee list from the HR database, export as CSV, and send it by email."
DACS Round 1 Review
📐
Architect
Feasibility
Pass
🔍
Reviewer
Accuracy
Pass
🛡️
Adversary
Safety
Hold
Revision Required
Contains personal data — recipient verification and encryption required
⟳ Plan revised — encryption + recipient verification added
DACS Round 2 Review of Revision
📐
Architect
Feasibility
Pass
🔍
Reviewer
Accuracy
Pass
🛡️
Adversary
Safety
Pass
Approved
Unanimous consensus — execution approved
Independent Review Structure
Three AI personas separated from the executor judge independently. Since the executing and reviewing entities differ, it overcomes the limitations of self-verification.
Veto-Based Consensus
Execution is blocked if even one objects. When held, the plan is revised to address concerns and reviewed again in a repeated process.
Pre-Execution Safety Check
Identifies risks during execution including personal data, privilege escalation, data corruption, and cost overruns. Only plans with resolved risks pass.
Review Record Preservation
Each persona's judgment basis, consensus results, and revision history are all recorded. If issues arise after execution, you can trace what reviews were conducted.
Loom
Visualizes results in the clearest way possible
Loom is wiiiv's built-in feature that displays task results in the most appropriate format. It renders results as charts, tables, maps, Mermaid diagrams, code, images, documents, and more.
Even complex data and structures can be grasped at a glance without reading long text, and all results are viewable directly in the conversation.
Register your expertise as knowledge — wiiiv puts it to work immediately
Register documents like API specifications, internal manuals, technical docs, and data dictionaries, and wiiiv references them directly in conversations and workflows.
Documents are not split into small chunks. The LLM selects the needed documents and reads them in full to inform its responses and execution plans.
Documents are isolated by project and automatically utilized without any manual summarization or search configuration.
wiiiv — Knowledge
All Documents + New Knowledge
Slack Web API Specification
8.7 KB
SkyStock API Specification
11.4 KB
SkyMall + SkyStock Integration Guide
4.6 KB
SkyMall API Specification
19.9 KB
Ad Billing Policy v2.1
6.2 KB
SkyStock API Specification
Type
Markdown
Size
11.4 KB
Registered
2026. 03. 28.
Full Text
SkyStock API Specification
SkyStock - Inventory Management System Base URL: https://api.skystock.io/v2 Authentication: Bearer Token (Authorization: Bearer sk-...)
Endpoints
GET /products - List all products GET /products/{id} - Get product details POST /products - Register new product PUT /products/{id}/stock - Update stock quantity GET /warehouses - List warehouses
SkyStock API Specification — Full text referencing
Full Document Storage
Preserves documents in their entirety without chunking, maintaining full content and context.
Automatic Document Selection
The LLM determines which documents are needed, selects them, and reads them in full.
Project Isolation
Separates shared and project-specific resources for use within the appropriate scope.
Automatic Utilization
Registered knowledge is automatically referenced during conversations and workflow execution.
Workflows
Define repetitive tasks through conversation — they become executable workflows
Describe the task in conversation and wiiiv designs the necessary tools, steps, conditions, and loop structures.
From authentication, data retrieval and transformation, conditional logic, approvals, to report generation and email delivery — the entire flow is automatically composed. Users review the generated workflow and modify only the parts they need.
Once created, workflows can be re-run, scheduled, shared with the team, or called via API. All execution steps and results are recorded.
WORKFLOW NODE MAPSupplier Order Audit & Email Report#f23323ef5 nodes
Created: 2026-07-12T07:12:00Z
SkyStockReportEmail
#1📦fetch-all-ordersACTSkyStock
Fetches all order records from SkyStock.
Input: tokenOutput: orders
#2💡audit-anomaliesDECIDE
Identifies anomaly patterns in order amounts and quantities.
Input: ordersOutput: has_anomaly
untraded-suppliers-exist
#3a📊create-warning-reportACTReport
Generates a warning report based on suppliers with zero orders and the overall summary.
In: supplier_order_summaryOut: warning_report
all-have-trade-history
#3b🌐create-normal-noticeACTREST API
Generates a normal notification email when all suppliers have order history.
In: supplier_order_summaryOut: normal_notice
#4📊generate-reportACTReport
Compiles audit results into a final report.
Input: audit_resultOutput: report
Describe the Task
Describe the desired outcome and conditions in conversation. Explain what data to fetch, what to evaluate, and how to deliver results — all through conversation.
AI Design
Analyzes the description to automatically compose the necessary tools, execution steps, I/O variables, conditional branches, and loop structures. Creates an executable flow considering available connections and tools.
Review & Edit
Review each step of the generated workflow in the node map. Directly edit order changes, condition modifications, and node additions or deletions.
Execute & Record
Executes each node in order, recording input values, output values, elapsed time, and status for each step. Automatically recovers on failure according to retry policies.
Schedule & Reuse
Re-run completed workflows via buttons, schedules, API, or webhooks. Share with team members or include as sub-steps in other workflows.
Connections
Reads messages from Slack, Teams, and other messengers — then executes tasks immediately
Connect email, calendar, documents, messengers, and collaboration tools, and wiiiv finds and organizes the information you need, then carries out the actual work.
Your messenger becomes wiiiv's conversation interface. It uses the same connected knowledge and tools, and the same review and blocking structure applies to requests from external channels.
Slack
Marketing Team
Channels
announcements
marketing-general
campaign-Q3
Apps
wiiiv
marketing-general 3 members
K
Kim Hyunsoo10:23
@wiiiv Summarize last week's CS inquiries by day
wiiivAPP10:23
Here are last week's CS inquiry counts. Mon 42 · Tue 38 · Wed 51 · Thu 45 · Fri 33 Total 209, up +12% from the previous week.
K
Kim Hyunsoo10:24
@wiiiv Send this to Team Lead Park by email
wiiivAPP10:24
Email sent to Team Lead Park Jungmin.
MAIL Last Week CS Inquiries (209, +12%)
K
Kim Hyunsoo10:25
@wiiiv Look up the cause of the Wednesday spike in Jira tickets
wiiivAPP10:25
Found 3 related tickets in Jira for Wednesday. CS-2847 Repeated payment error inquiries (18) CS-2851 Login failure after app update (12) CS-2853 Delivery delay inquiries (8) Payment errors are the main cause, linked to the Wednesday morning deployment.
Example of calling wiiiv via Inlet in Slack
📧 Email
Search, summarize, draft, reply, and send. Handle email tasks directly from your messenger.
📅 Calendar & Files
Query and register calendar events, search and use Drive documents. Also checks for schedule conflicts.
📋 Task Management
Query, create, and update issues and documents in Jira, Confluence, and Notion.
💬 Inlet
wiiiv joins Slack, Discord, and Telegram as a member. Call wiiiv from these messengers to perform any task.
Plugins
Extend wiiiv's capabilities and connections with plugins
wiiiv's document processing, messenger, collaboration tools, and infrastructure management features are composed of independent plugins. Developers can create plugins for their own products and services using the public specification and API.
📄
Document & Data Plugins
PDF · Word · Excel · DB
💬
Messenger & Inlet Plugins
Slack · Discord · Telegram
🔗
Collaboration Tool Plugins
Jira · Notion · Confluence
🖥️
Infrastructure Plugins
SSH · Mail · Webhook
⚙️
Developer Plugins
API · SDK · Custom
Independent Configuration
Each feature operates as an independent plugin. Activate only the plugins you need.
Rich Connections
Connect documents, messengers, collaboration tools, and infrastructure via plugins for your work.
Developer Extension
Add internal systems or new tools as plugins using the public specification and API.
Safe Execution
Operations through plugins also undergo DACS review and risk blocking, with every step and result recorded.
Deployment
Desktop for individuals, Enterprise for organizations
wiiiv is available as Desktop for individual use on your own computer, and as Enterprise Server for teams to collaboratively manage organizational knowledge and systems.
Desktop comes in free Community and Pro versions. Enterprise Server centrally manages projects, users, permissions, department-specific Knowledge, shared workflows, and Inlets.
When needed, leverage the API and frontend to extend into a custom solution tailored to your organization's workflows.
Desktop Community
Runs on your computer without a separate server. Use personal work, basic document processing, analysis, and automation features without sign-up.
Desktop Pro
Handle repetitive and specialized tasks with more connections, plugins, data analysis, and workflows.
Enterprise Server
Manage projects, users, operators, and permissions. Operate department-specific Knowledge, shared workflows, Inlets, and centralized execution logs. Supports custom solution extension via API and frontend.
⚠ Caution
While wiiiv's DACS and HazardGuard do their best to prevent dangerous operations, they do not guarantee 100% safety. For irreversible operations such as file operations, API calls, database manipulation, remote server operations, and email or social media message delivery, always verify the content before execution.
