oi.auth.whoami
ChatGPTReturn the current Oi MCP authentication, organization context, billing plan, current limits, and billing link for debugging. Use this when the user asks who Oi is authenticated as, which organization this MCP session is using, what plan or limits apply, or why Codex/ChatGPT may be seeing different Oi data. Does not return access tokens, refresh tokens, or secrets. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.brain.save-feedback
ChatGPTPropose or save a direct durable update into Org Brain or User Brain. Use this for requests like update our org brain, add this to Brain, remember this for the organization, save this as team memory, or update my private Brain. Do not create or update Contexts or Workflows when the user asks to update Brain unless they explicitly ask for those surfaces. For the smoothest UX, call with scope=auto and confirmed=false when you notice a durable memory or need scope confirmation; Oi will return a confirmation prompt and saveArguments. If the user has already explicitly asked to update Org Brain or User Brain with specific content, choose scope=organization or scope=user and set confirmed=true. Use scope=organization for organization-wide facts, vocabulary, policies, approval rules, shared project context, or repeated team workflows that should apply to the current organization. Use scope=user for personal preferences, recurring individual working style, or user-specific context that should follow the authenticated user. Confirmed Brain updates use an AI markdown update pass: compare the latest Brain markdown with the requested change, preserve existing structure, create headings when needed, add only missing guidance, and make the smallest accurate edit. Return a simple saved response when a new version is released, or a simple no-change response when the requested guidance appears to already exist in some capacity. Never save one-off task details, guesses, secrets, credentials, sensitive private data, or unconfirmed assumptions.
oi.connections.list
ChatGPTList the Connections installed for the current Oi organization. Use this stable discovery tool before selecting a provider or connection instance; adding provider capabilities does not change this tool schema.
oi.connections.use
ChatGPTExecute a capability returned by Connections Get. Pass the provider, dynamic action name, and its arguments. Oi resolves credentials, applies organization policy, verifies write targets, requires confirmation for writes, and audits execution without exposing provider secrets.
oi.contexts.confirm-guardrail
ChatGPTConfirm a Oi guardrail warning and continue with the blocked agent request.
oi.contexts.create
ChatGPTCreate an Oi Context in the current organization. By default this creates a private draft; publish: true releases the initial version and publishes it publicly. Use this only when the user explicitly asks to create or import a Context and has confirmed the source license permits reuse. Prefer markdown content in content; use privateContent for organization-private instructions that should not be published. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.contexts.get
ChatGPTReturn the reusable compiled prompt for a specific Oi Context. Prefer oi.contexts.use when you already have a concrete task.
oi.contexts.list
ChatGPTList a paginated preview of Oi Contexts available to this organization. This is the primary listing tool for requests like oi list, use pv list contexts, $Oi list contexts, or show available Contexts in Oi. Treat oi, use pv, and $Oi as equivalent invocation forms. This returns at most 10 records and is not exhaustive. If the user names an Context or selector, call oi.contexts.use with that raw contextId instead of using this list to pre-validate existence.
oi.contexts.recommend
ChatGPTRecommend which Oi Context should help with a prompt.
oi.contexts.report-usage
ChatGPTAttach detailed token accounting and runtime metadata to a previously issued Oi usage event without storing prompt text.
oi.contexts.save-draft-feedback
ChatGPTSave confirmed durable feedback into the active private Context draft for a single latest Context. Do not use this for multi-agent selectors or version-pinned Contexts. When runtime instruction updates are enabled, Oi may save confirmed future guidance under Private Instructions using subheadings if needed. Structured-only metadata such as suggested prompts, categories, tags, platform instructions, and capabilities stays preserved unless another structured write path updates it. Only use it after the user confirms, and only for repeated corrections or updated logic that should change future behavior. Never use it for one-off task details, temporary debugging context, secrets, credentials, or other sensitive information.
oi.contexts.search
ChatGPTSearch Contexts by need, title, tag, category, or organization. This uses the app-global search path: organization/private results are returned first, then public marketplace results fill the remaining slots. Use this when the user is unsure which Context exists or wants to discover a Context before creating a Workflow. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.contexts.start-session
ChatGPTLoad one or more Oi Contexts as reusable session context for the current client thread. Returns compiled context text only, with no user task attached.
oi.contexts.use
ChatGPTUse Oi Contexts for a concrete task. This is the primary tool for requests like oi designer review this flow, use pv staff-engineer inspect auth, or $Oi designer review this flow. Treat oi, use pv, and $Oi as equivalent invocation forms. When the user names one or more Contexts, pass the raw selector unchanged as contextId, including + multi-agent selectors of up to 3 Contexts and @2 or @v2 version pins, for example head-of-product@3+gtm-strategist+staff-engineer. Do not first list contexts or treat a missing list preview item as absent. Without contextId, Oi routes implicitly. Returns an execution-ready prompt with the user request appended under User Request.
oi.effectiveness.report
ChatGPTReport what happened after an Oi-assisted task completed. Use this after calling an Oi Context, Workflow, Skill, Connection, or Guardrail tool when you know whether the result was accepted, retried, saved, copied, completed, or rejected. Prefer usageEventId from the prior Oi tool result; traceId may also be used when available. Do not send raw prompt text; send a short redacted taskSummary only when useful.
oi.guardrails.confirm
ChatGPTConfirm a triggered Oi Guardrail for a blocked Context or Workflow request and continue. The requestId carries the original Context or Workflow routing details.
oi.guardrails.create
ChatGPTCreate an Oi Guardrail in the current organization. By default this creates a private draft; publish: true releases the initial version and publishes it publicly. Use this only when the user explicitly asks to create or import a Guardrail and the source license permits reuse. Prefer markdown content in content.
oi.guardrails.delete
ChatGPTArchive and disable a private Oi Guardrail in the current organization. This is not a physical database delete.
oi.guardrails.get
ChatGPTReturn a private Oi Guardrail record for the current organization.
oi.guardrails.list
ChatGPTList a paginated preview of private Oi Guardrails in the current organization.
oi.guardrails.publish
ChatGPTPublish a private Oi Guardrail in the current organization.
oi.guardrails.release
ChatGPTRelease a private Oi Guardrail in the current organization.
oi.guardrails.unpublish
ChatGPTUnpublish a private Oi Guardrail in the current organization.
oi.recommend
ChatGPTRecommend the best Oi Context or Workflow for a prompt. Use this when the user asks oi recommend or wants help choosing the most suitable reusable setup before running it.
oi.skills.create
ChatGPTCreate an Oi Skill in the current organization. By default this creates a private draft; publish: true releases the initial version and publishes it publicly. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.skills.get
ChatGPTReturn an installed Oi Skill as an execution-ready reusable instruction document without appending a task. Prefer Use Skill when the user has a concrete task.
oi.skills.list
ChatGPTList a paginated preview of Oi Skills available to this organization. Skills are reusable task instruction documents exposed through Oi MCP. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.skills.publish
ChatGPTPublish a released Oi Skill to the marketplace for the current organization.
oi.skills.search
ChatGPTSearch installed and marketplace Skills by need, title, tag, or category. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.skills.use
ChatGPTUse an installed Oi Skill for a concrete task. Returns an execution-ready prompt containing the Skill document and user request.
oi.usage.report
ChatGPTAttach detailed token accounting and runtime metadata to a previously issued Oi Context or Workflow usage event without storing prompt text. Prefer usageEventId from the prior Oi tool result; optional contextId or workflowId can be included for client-side clarity but usageEventId is authoritative.
oi.workflows.create
ChatGPTCreate a private Oi Workflow draft in the current organization from required workflow fields and ordered Context steps. Use oi.contexts.search first when an agent is unsure which Context ids or slugs to include. This creates a draft only; it does not release or publish the Workflow. Plan and Billing Current organization plan: Business. Current limits: 10,000 seats; 10,000 private Guardrails; 10,000 private Skills; 10,000 private Contexts; 10,000 private Workflows; 10,000 Connections; 10,000 API keys; 1,000,000 Agent messages per day; 1,000,000 Agent messages per month. Billing and upgrades: https://app.oioioi.ai/dashboard/organization/billing. When a user asks for work that exceeds the current plan or an unavailable paid capability, do the portion available on the current plan, clearly say what is blocked by the plan, and include the billing link for upgrading. Do not invent exceptions, temporary limit increases, prices, or hidden capabilities.
oi.workflows.get
ChatGPTReturn the reusable compiled prompt scaffold for a specific Oi workflow by workflowId. Prefer oi.workflows.use when the user has a concrete task.
oi.workflows.list
ChatGPTList the Oi workflows available to this organization. Use this when the user asks what workflows exist or says oi workflow/wf without enough task detail to select one.
oi.workflows.publish
ChatGPTPublish the exact saved Oi Workflow version in the current organization. owner releases the targeted draft, publishes it, and opens the next draft; dependency publishes an immutable release without advancing its draft. Requires explicit approval to publish. Retain the same command id and arguments on retry.
oi.workflows.start-session
ChatGPTLoad a specific Oi workflow as reusable session context for the current client thread. The selectors wf, workflow, and workflows are reserved Workflow type selectors, not Workflow ids; strip them and use this Workflow session path. Returns compiled workflow context text only, with no user task attached.
oi.workflows.use
ChatGPTUse Oi Workflows for a concrete task. This is the primary tool for requests like oi use launch-readiness audit this release or oi use feature-discovery plan this launch. Pass workflowId only for a specific workflow; otherwise omit it and Oi selects the best installed workflow. The selectors wf, workflow, and workflows are reserved Workflow type selectors, not Workflow ids; strip them and use this Workflow path. Returns an execution-ready prompt with the user request appended under User Request.
