DocsBot

Copy a skill between authorized bots Copies one skill from a different authorized source bot into the destination bot, either within the same workspace or across workspaces the caller can access. Destination teamId/botId select where to create the copy; sourceTeamId/sourceBotId/sourceSkillId select the existing source. Use list_importable_skills to discover source IDs; no personal profile identifiers or credentials are requested. Copies a published source package into both draft and published storage; unpublished skills copy into draft storage only. Copies are independent and disabled. Clears saved settings, secrets, and runtime history. Conflicting IDs get numbered names without overwriting existing skills. Published copies can be configured and enabled without republishing; draft-only copies must be published first. MCP returns only the new copied skill resource IDs, names, status, enablement flags and publication timestamp; internal storage paths, configuration, credentials and diagnostics are omitted. Use get_skill_settings to inspect copied configuration. Requires: canUserManageBotSettings and Standard plan or higher on destination; live team membership and canUserViewBot on source. Effect: write. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

copy_skill_from_bot

Count bot sources by status Returns source counts grouped by status for dashboard summaries and ingest monitoring. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

count_sources

Count indexed source retriever tags Returns counts for configured retriever tags on the bot, useful before filtering source lists or search by tag. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

count_source_tags

Create a bot in a team Creates a pending bot, increments team botCount in a transaction, creates vector storage when vectorDatabase is weaviate, optionally queues an async copy from another bot, applies best-effort brand enrichment from brandSource/siteURL/brandAnalysis, clears Cloudflare cache, and returns the created bot. When theme is omitted, new bots are saved with theme=auto. For business bots created through Admin MCP, include brandSource.siteURL whenever a public website is known, or brandSource.analysis when reusing a prior analysis response. Omit brandSource only when no public website or reusable brand analysis is available. Requires: Requires bot creation or deletion permission via existing RBAC.. Effect: write. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

create_bot

Create a bot webhook Creates an outgoing webhook for one or more events. The target URL is normalized and validated; unsupported events and invalid expiration dates are rejected. Requires: Requires an authenticated DocsBot dashboard session for the requested team. The handler preserves existing team and bot RBAC; the Admin MCP token only identifies the dashboard user. Webhook operations require canUserManageIntegrations for the bot. Team integration connect/configure routes require canUserModifyTeam or canUserManageIntegrations as noted.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

create_webhook

Create a signed upload URL for a bot widget image Returns a 30 minute, write-only Google Cloud Storage signed URL for a widget branding image plus the dashboard-style storage path and public CDN URL to save into bot logo, icon, or botIcon. This is for bot appearance assets, not knowledge source files. Requires: Requires bot settings manager via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

create_bot_image_upload_url

Create a signed upload URL for a source file Returns a 30 minute, write-only Google Cloud Storage signed URL plus the pending storage path to pass as file when creating a file-backed source. This route only signs an upload target; the MCP execute tool does not stream local bytes to that signed URL. Requires: Requires source modification permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

create_source_upload_url

Create a team for the delegated user Creates a new Firestore team owned by the current user, sets users/{uid}.currentTeam to the new team, and returns the created team with its resolved plan. If name is omitted, the API derives a unique "<user name or email prefix>'s Team" name. MCP omits credential values/placeholders, partial credential suffixes and internal provider endpoints from workspace responses. Membership responses omit avatars; role and configured/status fields remain available. Requires: Requires authenticated DocsBot delegated user access.. Effect: write. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

create_team

Create and queue ingestion for a bot knowledge source Add training content to the single bot identified by teamId and botId; requires source-modification permission and applicable plan limits. Creates a source record, moves supplied pending uploads into that source folder, updates source counts, and queues asynchronous ingestion. Depending on source type, ingestion fetches the supplied website or URL, processes uploaded files, or syncs the selected connected external provider. A qa request can merge FAQs into the existing Q&A source and queue its reingestion instead of creating a second source. The response acknowledges creation or merging, not completed indexing. Inspect existing sources first and check the returned source status before retrying an uncertain request. Provider credentials are entered in the dashboard, not accepted by this tool. MCP omits internal storage paths and provider linkage from the returned source record. Source IDs, content details, status and diagnostics remain available; ingestion inputs are unchanged. Requires: Requires source modification permission via existing RBAC.. Effect: write. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

create_source

Delete a bot Deletes the selected bot and its stored data, revokes its CRM connections, removes Truto integrated accounts, cleans its vector namespace or tenant, and clears Cloudflare cache. Some external cleanup is best effort; success does not guarantee every external resource was removed. Research jobs use their dedicated cancellation route. Requires: Requires bot creation or deletion permission via existing RBAC.. Effect: delete. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

delete_bot

Delete a bot webhook Removes one webhook from the bot configuration. Requires: Requires an authenticated DocsBot dashboard session for the requested team. The handler preserves existing team and bot RBAC; the Admin MCP token only identifies the dashboard user. Webhook operations require canUserManageIntegrations for the bot. Team integration connect/configure routes require canUserModifyTeam or canUserManageIntegrations as noted.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

delete_webhook

Delete a knowledge source Deletes a ready or failed source and clears matching team lastError. Queues vector cleanup, deletes source storage files and, for connector-backed sources, attempts to delete the external Truto integrated account. External cleanup may be best effort. Sources that are still pending or indexing must finish before deletion. Requires: Requires source modification permission via existing RBAC.. Effect: delete. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

delete_source

Delete one conversation Hard-deletes a conversation document from the bot conversations collection. Requires: Requires bot editor permission via existing RBAC.. Effect: delete. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

delete_conversation

Delete one lead Hard-deletes a lead document from the bot leads collection. Requires: Requires bot editor permission via existing RBAC.. Effect: delete. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

delete_lead

Disconnect OAuth for an external MCP server Deletes DocsBot stored OAuth registration and token state for the normalized external MCP server URL. It does not revoke the client registration at the upstream authorization server. Requires: Requires an authenticated DocsBot dashboard session for the requested team. The handler preserves existing team and bot RBAC; the Admin MCP token only identifies the dashboard user. Requires canUserManageBotSettings for the bot, except read endpoints that permit normal bot access.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

disconnect_external_mcp_oauth

Disconnect one Slack workspace Removes a Slack workspace from the Slack integration document and from team.slackWorkspaceIds. MCP omits credential values/placeholders, partial credential suffixes and internal provider endpoints from workspace responses. Membership responses omit avatars; role and configured/status fields remain available. Requires: canUserModifyTeam or super admin. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

disconnect_slack_workspace

Disconnect the Twilio voice account Requires team admin or owner permission. Removes stored Twilio connection credentials. The request fails while any phone numbers remain assigned. MCP omits credential values/placeholders, partial credential suffixes and internal provider endpoints from workspace responses. Membership responses omit avatars; role and configured/status fields remain available. Requires: Requires team admin or owner permission via existing RBAC.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

disconnect_twilio_account

Discover tools on an external MCP server Validates the outbound MCP server URL, uses saved custom headers and OAuth tokens for the matching bot MCP server, refreshes expired OAuth tokens when possible, probes OAuth metadata when auth is needed, then runs MCP initialize/tools-list discovery. Requires: Requires an authenticated DocsBot dashboard session for the requested team. The handler preserves existing team and bot RBAC; the Admin MCP token only identifies the dashboard user. Requires canUserManageBotSettings for the bot, except read endpoints that permit normal bot access.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

discover_external_mcp_tools

Draft a title for a batch of uploaded source files Generates a concise title from uploaded file names and existing source titles. One file or missing OpenAI configuration returns a deterministic fallback title. Requires: Requires source modification permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

draft_source_title

Draft an improved bot prompt Uses OpenAI to generate a system prompt from an input task or existing prompt. activeTab=agent uses the agent-specific meta prompt that emphasizes search_documentation. This does not persist the generated prompt. Requires: Requires bot editor permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

draft_bot_prompt

Draft custom CTA button action metadata Uses OpenAI to draft custom button action fields from a natural language description. It validates plan access and action slot availability, but only returns draft metadata; it does not save the button. Requires: Requires bot settings manager via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

draft_custom_button

Draft remote MCP server metadata Uses OpenAI and low-context web search to draft an internal serverLabel and serverDescription for a remote HTTPS MCP server. enabledTools can constrain the generated description. It validates plan access and action slot availability, but does not save the connector. Requires: Requires bot settings manager via existing RBAC.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

draft_mcp_server

Export a Q&A source to CSV Builds a CSV from a qa source faqs array, stores a unique, create-only download in the user export folder, and returns a 7 day signed URL with a friendly filename. Source records and earlier downloads are unchanged. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

export_qa_source

Export a bot question log as CSV Reads bot question history for a required date range, applies optional filters, and returns a CSV download URL that expires after seven days. The temporary download uses a unique, create-only storage object and a friendly download filename; repeated exports do not replace earlier files or modify question records. Requires: canUserExportBotLogs and Personal plan or higher. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

export_question_log

Export bot leads to CSV Reads leads in a required date range and returns a CSV download URL that expires after seven days. The temporary download uses a unique, create-only storage object and a friendly download filename; repeated exports do not replace earlier files or modify lead records. Metadata keys become CSV columns; name and email are ordered first when present. Requires: Requires bot log export permission via existing RBAC.. Effect: write. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

export_leads

Get a bot monthly report Lists available report month document IDs and returns the requested report. If month is missing or unavailable, the route falls back to the first available report ID returned by Firestore. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_monthly_report

Get a bot retrieval OpenAPI document Returns a generated OpenAPI 3.1 document for the bot retrieval search endpoint on api.docsbot.ai. The schema includes retriever tag enum fields when the bot has retrieverTags. Private bots include bearerAuth security scheme metadata. Requires: Requires access to the target team via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_bot_retrieval_openapi

Get a signed download URL for a source file Returns a 7 day signed URL for a downloadable source. For urls sources with indexedUrls but no stored file, this route generates a unique, create-only indexed-urls.csv download. It does not change the source record or replace an existing file. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_source_download_url

Get a signed download URL for one file in a multi-file source Resolves an existing file by name within a multi-file source and returns a 7 day signed URL plus the resolved filename. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_source_file_download_url

Get bot aggregate stats Returns aggregate bot stats either for a timeDelta window or for a startDate/endDate range. Date ranges are clamped to at most 365 days; invalid or missing timeDelta defaults to 30 days. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_bot_stats

Get bot settings Returns sanitized bot settings after bot-level view permission checks. Use the dedicated research list and job tools for research jobs. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_bot

Get direct SIP routing for a bot Requires access to the target team and bot. Returns the direct SIP destinations plus Twilio-managed phone routes without changing configuration. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_sip_routing

Get one bot webhook Returns one normalized outgoing webhook by id. Requires: Requires an authenticated DocsBot dashboard session for the requested team. The handler preserves existing team and bot RBAC; the Admin MCP token only identifies the dashboard user. Webhook operations require canUserManageIntegrations for the bot. Team integration connect/configure routes require canUserModifyTeam or canUserManageIntegrations as noted.. Effect: integration. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_webhook

Get one conversation transcript Fetches a single conversation document after verifying team access, bot existence, and bot-level view permission. MCP omits known structured visitor identifiers and contact details; custom metadata is limited to bounded topic/language/sentiment labels and debug objects to numeric timing/token counts; requested transcript and source text can still contain personal information. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_conversation

Get one knowledge source Fetches one source. Truto-backed sources may include trutoSelected if the route can load the selected connector files. MCP omits internal storage paths and provider linkage from the returned source record. Source IDs, content details, status and diagnostics remain available; ingestion inputs are unchanged. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_source

Get one research job Fetches one research job and normalizes createdAt, completedAt, and cancelledAt/deletedAt timestamp fields when possible. Requires: Requires bot view permission via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_research_job

Get sanitized team settings Returns one team after userTeamCheck, removes keys whose names start with "stripe", and adds the resolved plan object. MCP omits credential values/placeholders, partial credential suffixes and internal provider endpoints from workspace responses. Membership responses omit avatars; role and configured/status fields remain available. Requires: Requires access to the target team via existing RBAC.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_team

Get the advertised schema and metadata for one fixed public DocsBot Admin tool name.

get_tool_schema

Import a library skill into a bot Copies the library package into the target bot draft and published package storage and creates a new skill record with status=draft. The imported skill starts disabled for the bot, widget and voice; copying the published package does not enable execution. Requires Standard plan or higher. MCP returns resource IDs, display/state summaries and copy counts only; publisher identity, original team/bot and storage paths, credentials and raw diagnostics are omitted. Use get_skill_settings to inspect the imported skill configuration. Requires: canUserManageBotSettings; Standard plan or higher. Effect: write. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

import_library_skill

Inspect skill draft and published packages Reads supported package files from R2 for the draft package and, when publishedAt is set, the published/current package. Only SKILL.md, package.json, bundle output, scripts, references, and assets are returned. Requires: Requires an authenticated DocsBot dashboard session for the requested team. The handler preserves existing team and bot RBAC; the Admin MCP token only identifies the dashboard user.. Effect: read. DocsBot Admin API documentation: https://docsbot.ai/documentation/developer/admin-api-overview.

get_skill_package