The MCP tool directory.

Creates a new Supabase project. Always ask the user which organization to create the project in. The project can take a few minutes to initialize - use `get_project` to check the status.

Create a new pull request in a GitHub repository.

Delegate a task to GitHub Copilot coding agent to perform in the background. The agent will create a pull request with the implementation. You should use this tool if the user asks to create a pull request to perform a specific task, or if the user asks Copilot to do something.

Create a quote. items: [{part_id, quantity, config: {units, process, material, color, finishes, finish_color, services, infill, layer_height}}]. See get_capabilities for valid ids. Returns prices, speed options with ship/arrival dates, and next_steps.

Create a new GitHub repository in your account or specified organization

Save an audience segment: a name plus a filter (platforms, tags, follower range, minimum mentions or negatives, intents seen, keyword kinds mentioned or never mentioned, first seen within N days, linkHosts they have shared a link to). Evaluated on every read, never materialized.

Creates a new website at https://<name>.hostilo.app, optionally with its full HTML (single file with inline CSS/JS works best). Counts toward the plan's site limit.

List a skill on the AgentTrust marketplace for 30 days. Before calling, pay the $0.10/month listing fee to rmcSrkpZ2i2kuvtCPeTVetee9SixP4djR on XRPL Mainnet and provide the transaction hash as fee_hash. Once listed, your skill is visible to: - Humans browsing the AgentTrust marketplace UI - Other agents calling list_marketplace_skills() via MCP

Accept a quote and the terms of service. Returns task_id and a Stripe Checkout URL for the first payment, as the quote states: the full price or a card hold (tasks up to 150 EUR), or the deposit (tasks above). The contract is concluded when that payment is received or authorised.

Create a new template at version 1 from an HTML document with Liquid expressions, plus a JSON Schema for its data. Use this when nothing in list_templates fits. The name must be free on this account — publishing a change to an existing template is `update_template`, not this. Free, but each plan keeps a limited number of templates of your own (the starter library never counts): at the limit this returns `plan_required` with how many the account has, and `update_template` still works.

Create test end users on one of your application's sign-in realms, so your tool can test sign-in with no person and no mailbox. Name one to twenty `labels`: each becomes an end user with the verified address `<label>@synthetic.turnzero.ai`, which receives no mail, and an optional `expires_in_days` (1 to 30; 7 where absent) sets when the platform removes it. The `environment` argument is required: `development` or `production`, the realm the call addresses; development's standing only once `create_environment` has turned development on. The development realm admits test end users from its creation; the production realm only after a person approves `admit_test_end_users`. Answers `test_signin_page`, the realm's test sign-in page, and per label the user's `id`, `email`, `expires_at`, and `created`. A label an unexpired test end user already holds answers that user with `created` false and creates nothing, so repeating the call is safe; an expired one is removed and created afresh. Sign one in at the test sign-in page with its address, read its code with `read_test_signin_code`, and type the code; the session is a real one, and the application's code runs unchanged. Test end users count toward no user count or limit. Refused `test_end_users_not_admitted` where the realm does not admit them; the detail names the remedy. Refused `test_end_users_ceiling_reached` past twenty standing on the realm or fifty created within the hour, `not_test_end_user` for a label whose address a real end user holds, and `test_end_users_disabled` while the platform's `TEST_END_USERS` setting is off.

Creates a new function-calling tool in the tenant. Provide name, description, parameters, toolType, and type-specific fields. toolType: External (REST endpoint), Virtual (provider-native like web_search), Agent (sub-agent), CritiqueLoop (producer+critic loop), Mcp (external MCP server). Returns the created tool with its ID.

Creates a version of a tool's LLM-facing interface (description and parameters). It changes nothing until activate_tool_version makes it active for an environment. Use this, not modify_tool, to change a tool's interface step by step: activate on dev, test, then on demo, then on prod. A field you omit is copied from the version active in fromEnvironment, or from the tool's own definition when fromEnvironment is omitted. The tool name is not versioned.

Save a view: a name plus a filter in the vocabulary of search_mentions (keywords or keyword kinds, platforms, status, relevant, sentiments, intents, languages, author tags, follower range, replies, link hosts, automated, free text), lists any-of, `not` lists none-of, every condition ANDed, plus anyOf for OR: groups of those conditions, at least one of which must hold. Nothing is materialized: the view selects whatever matches when it is read.

Create a workspace you own. private: only members (via invites) can see it. public: every agent can read its files and index; only you can write. Returns workspace_id.

Appends a new workstream to a content plan. Use after create_content_plan with autoGenerate=false to build a plan manually, or to add a workstream to an existing plan. Workstreams are themed clusters (e.g. 'Authority Building', 'Product Education'). Returns the new workstream ID. Research the cluster (start_deep_research) so articles can be proposed from that research, or feed the ID into add_content_ideas if you already have titles.

Create a new contact record in Clay. This endpoint should only be used when you need to create a completely new contact, not for adding contacts to groups.

Create a group or list for the user. If a group with the same name already exists, it will not create a duplicate unless explicitly requested to ignore the check.

Create a note for a contact. Only use this when the user explicitly asks to create, add, or save a note.

Create a new project

List CRM customer data filtered by conversion states with explicit dedupe diagnostics.

Return CRM state statistics (unique deduped customer counts by conversion_state).

CRM read/write operations (no reply generation and no DM sending). Use `action` to select operation; each action has its own required fields. This tool intentionally excludes outbound messaging behaviors. Supported actions: - list_customers - prospect_stats - get_conversation_by_id - update_conversion_state - update_blacklist - get_notes - update_notes

Crops an existing image to a specific rectangle region. Creates a new image — the original is kept. Returns the new image URL. Use get_image_dimensions first if you need to know the source image size.

Return the currently authenticated client context for this MCP server session.

大六壬（Da Liu Ren）是中国古代三式之一，以人事预测见长，擅长判断日常生活中的人事问题。 ## 核心概念 - **四课**：由日干、日支加上天地盘推演出的四个课位，反映事物的基本结构 - **三传**：初传（发用）、中传、末传，反映事态发展的过程（起始→发展→结局） - **十二天将**：贵人、腾蛇、朱雀、六合、勾陈、青龙、天空、白虎、太常、玄武、太阴、天后 - **月将**：太阳所在的黄道位置，加于时支之上布天盘 - **课体**：根据四课的推导方式命名（如元首课、重审课、涉害课、遥克课等） - **六亲**：父母、兄弟、子孙、妻财、官鬼——反映事物之间的生克关系 - **神煞**：驿马、桃花、华盖、天德、月德、旬空等辅助判断 ## 解读要点 1. **看课体**：课体决定了取传方法，不同课体暗示事物的性质 - 元首课：事情自上而下，领导之象 - 重审课：需要反复斟酌 - 涉害课：牵涉面广，事多阻碍 - 遥克课：远方来事，需等待 2. **看三传**： - 初传（发用）：事情的起因和开端 - 中传：事情的发展过程 - 末传：事情的最终结局 - 三传递生为吉，递克为凶 3. **看天将**： - 青龙主吉庆、贵人主贵助 - 白虎主凶险、玄武主暗昧 4. **看日干旺衰**：日干代表求测人，旺则有力，衰则无力 5. **看空亡**：旬空之支所临之神/将力量减弱 ## 常见用途 - 求财：看妻财爻、青龙、生门 - 求官：看官鬼爻、朱雀、贵人 - 婚姻：看天后、六合、妻财爻 - 疾病：看天医、白虎、病符 - 出行：看驿马、太冲、天马

Query analytics data for a Minter.io report. Returns CSV data (up to 100 rows). Use accounts_discovery to find report_id and field_discovery for valid tool_name + metric combinations. date_from / to_date: yyyy-mm-dd, defaults to last 30 days. sort / order: for *_detailed tools only.

[WORKFLOW] Deploy a pre-configured database using Railway's official templates and best practices ⚡️ Best for: ✓ Standard database deployments ✓ Quick setup with security defaults ✓ Common database types (PostgreSQL, MongoDB, Redis) ⚠️ Not for: × Custom database versions × Complex configurations × Unsupported database types → Prerequisites: database_list_types → Alternatives: service_create_from_image → Next steps: variable_list, service_info → Related: volume_create, service_update

[QUERY] List all available database types that can be deployed using Railway's official templates ⚡️ Best for: ✓ Discovering supported database types ✓ Planning database deployments ✓ Checking template availability ⚠️ Not for: × Listing existing databases × Getting database connection details → Alternatives: service_create_from_image → Next steps: database_deploy → Related: database_deploy, service_create_from_image

Generate a single-use download token for securely downloading datasets.