{"name":"ThinkPlatform Docs","description":"Learn to use ThinkPlatform, build integrations, and manage your organization.","url":"https://docs.thinkplatform.ai/","version":"1.0.0","protocolVersion":"0.3","preferredTransport":"HTTP+JSON","supportedInterfaces":[{"url":"https://docs.thinkplatform.ai/","protocolBinding":"HTTP+JSON","protocolVersion":"0.3"}],"provider":{"url":"https://docs.thinkplatform.ai/","organization":"ThinkPlatform Docs"},"documentationUrl":"https://docs.thinkplatform.ai/","capabilities":{"streaming":false,"pushNotifications":false},"defaultInputModes":["text/plain"],"defaultOutputModes":["text/plain"],"skills":[{"id":"thinkplatform-agent-cost-analysis","name":"thinkplatform-agent-cost-analysis","description":"Use when investigating, explaining, auditing, or reporting unusually high ThinkPlatform Agent or LLM costs with the tp CLI. Trigger for requests about expensive projects, model spend, scheduled Agent runs, token or cache efficiency, prompt/context bloat, repeated Blob/PDF/OCR/tool work, runaway tasks, retries, budget risk, or cost comparisons across projects, Agents, tasks, users, models, and time windows.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-agent-cost-analysis/skill.md"},{"id":"thinkplatform-agent-ops","name":"thinkplatform-agent-ops","description":"Use when creating, updating, auditing, or improving ThinkPlatform Agents, Agent teams, task workflows, workspaces, memories, system tools, remote tools, skills, and autonomous-agent operating patterns. Trigger for requests about Agent management, team design, model/tool selection, skill assignment, task inspection, memory cleanup, workspace preparation, or when an Agent should improve its own operating setup.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-agent-ops/skill.md"},{"id":"thinkplatform-automation-ops","name":"thinkplatform-automation-ops","description":"Use when designing, creating, auditing, or improving ThinkPlatform automations: scheduled checks, recurring Agent tasks, inbound Hook Apps, outbound Webhooks, follow-ups, monitoring loops, notification flows, and autonomous improvement proposals. Trigger for requests like 'run this every morning', 'monitor this and alert me', 'create a webhook', 'turn this repeated work into an automation', or when an Agent notices a repeated manual workflow that should become observable automation.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-automation-ops/skill.md"},{"id":"thinkplatform-autonomous-plant-ops","name":"thinkplatform-autonomous-plant-ops","description":"Use when coordinating end-to-end autonomous plant operations and improvement loops across ThinkPlatform: assess current plant/project state, detect issues or improvement opportunities, form hypotheses, select data/MLOps/edge/automation/agent/canvas work, execute or propose safe next actions, and close the loop with verification. Trigger for broad operations requests, autonomous improvement, plant optimization, recurring operational review, cross-domain incident triage, or 'what should the Agent improve next?' tasks.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-autonomous-plant-ops/skill.md"},{"id":"thinkplatform-business-workflows","name":"thinkplatform-business-workflows","description":"Use when building, updating, or validating ThinkPlatform business workflows and internal apps that combine Canvas pages, shared DB tables, Data Lake files, Code Flow actions, Dynamic UI, Module calls, and optional ML/Agent execution. Trigger for CSV/file processing apps, inspection records, approvals, report generation, operator workflows, data intake/review, and non-ML business processes that should run inside ThinkPlatform without a separate app deployment.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-business-workflows/skill.md"},{"id":"thinkplatform-canvas-interface","name":"thinkplatform-canvas-interface","description":"Use when creating, updating, auditing, or debugging ThinkPlatform Canvas/Home/Dynamic UI interfaces for users: dashboards, runbooks, review screens, operator panels, data/ML/edge status views, forms, embedded tables, Dynamic UI apps, Mermaid diagrams, and visualization pages. Trigger when the main deliverable is a human-facing interface, operational visualization, or Canvas-based workflow surface.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-canvas-interface/skill.md"},{"id":"thinkplatform-data-engineering","name":"thinkplatform-data-engineering","description":"Use when building or operating ThinkPlatform data engineering and analytics assets: Data Lake ingestion, file cleanup, shared DB tables, Databricks/Lakehouse schemas and Delta tables, SQL previews, feature tables, Gold analytics tables, data quality checks, and handoff datasets for MLOps or Canvas dashboards. Trigger for requests to analyze data, prepare assets, convert files to tables, build pipelines, create metrics, or improve data reliability.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-data-engineering/skill.md"},{"id":"thinkplatform-edge-ops","name":"thinkplatform-edge-ops","description":"Use when managing ThinkPlatform IoT Edge devices and edge deployments: device inventory, provisioning/setup scripts, module status, module restart, direct methods, module twin inspection/update, ACR image checks, IoT deployment creation/update, edge model rollout, readiness checks, troubleshooting, and rollback planning. Trigger for edge device, deployment, module, IoT Hub, or plant-site runtime operations.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-edge-ops/skill.md"},{"id":"thinkplatform-integrations-and-flows","name":"thinkplatform-integrations-and-flows","description":"Use when connecting ThinkPlatform to external systems or composing executable platform flows: registered Modules, MCP/OpenAPI tools, Code Flow, graph Flow creation/saving/running, Hook-triggered flows, API fallback via tp api/raw/rest, input/output schema design, integration smoke tests, and workflow wiring across Data Lake, DB, Canvas, Agents, MLOps, and Edge. Trigger for external API integration, Module invocation, Flow construction, or multi-step tool orchestration.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-integrations-and-flows/skill.md"},{"id":"thinkplatform-lakehouse-ai","name":"thinkplatform-lakehouse-ai","description":"Use when building, validating, or operating Databricks AI Functions enrichment through ThinkPlatform: discover function syntax and workspace availability, write SQL or Python/SQL notebooks, run small Serverless validations, create Lakeflow Jobs, inspect outputs, or wrap a stable AI operation in a project-catalog SQL UDF. Trigger for ai_summarize, ai_extract, ai_classify, ai_query, document parsing, translation, sentiment, semantic enrichment, and structured LLM output on Lakehouse data.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-lakehouse-ai/skill.md"},{"id":"thinkplatform-lakehouse-cost-analysis","name":"thinkplatform-lakehouse-cost-analysis","description":"Use when investigating, explaining, auditing, or reporting ThinkPlatform Lakehouse and Databricks cost through tp CLI or ThinkPlatform MCP system tools. Trigger for high DBU or Azure ActualCost, idle SQL Warehouses, frequent or duplicated SQL, cache misses, expensive or failed Jobs, All-Purpose compute, missing cost tags, project allocation, billing coverage, or before/after cost comparisons.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-lakehouse-cost-analysis/skill.md"},{"id":"thinkplatform-mlops","name":"thinkplatform-mlops","description":"Use when operating ThinkPlatform MLOps workflows: register data assets, create or update custom training bundles, submit and monitor training jobs, compare metrics, register model versions, validate deployability, deploy or update Azure ML online endpoints, invoke inference, replace deployed models, and iterate on model accuracy. Trigger for model creation, training, deployment, inference, accuracy improvement, or ML asset lifecycle requests.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-mlops/skill.md"},{"id":"thinkplatform-app-development","name":"thinkplatform-app-development","description":"Build, publish, update and manage container-backed ThinkPlatform web apps using the TP CLI and Python SDK, including publishing source from an Agent Workspace, displaying the app on a Canvas application page, and accessing TP Tables or Lakehouse as the signed-in user.","tags":[],"url":"https://docs.thinkplatform.ai/.well-known/agent-skills/thinkplatform-app-development/skill.md"}]}