OpenApps

AIsa MCP server — tools & history

AIsa MCP server: observed open status, tool schemas, linked apps and version history.

AIsa

Endpoint: https://tools.aisa.one/mcp

Observed status: open. Checked: 2026-10-12T05:06:02+00:00.

Linked apps

  • AIsa

    AIsa-plugins plugin for Cursor

Observed tools

  • batch_get_schema — Retrieve schemas for 1 to 20 exact tool names from search. Build arguments from successful schemas, then call batch_quote. Do not call batch_use yet. Reuse the session_id returned by Search in all subsequent tool calls for this task. If omitted, a new session is created.
  • batch_quote — Obtain AIsaServices cost quotes for 1 to 20 published tools without executing them. Same input as batch_use: exact tool names and schema-valid arguments. Never falls back to execution. The response reports auto_execute_eligible and auto_execute_threshold_micros_usd. auto_execute_eligible waives only the extra cost-confirmation step and does not override a user prohibition on execution, a narrower scope, or a tighter budget limit. Quotes are price observations, not authorization grants. Reuse the session_id returned by Search in all subsequent tool calls for this task. If omitted, a new session is created.
  • batch_use — Billable API calls can incur substantial charges. Never call before a successful batch_quote for the same tools and arguments. Execute only when that quote response reports auto_execute_eligible=true for these calls, or when explicit user authorization covers that cost and risk. auto_execute_eligible waives only the extra cost-confirmation step; it does not override a user prohibition on execution, a narrower scope, or a tighter budget limit. Re-quote if tools, arguments, or scope change. Execute 1 to 20 independent published-tool calls; never silently retry. A missing or failed quote is not free. Reuse the session_id returned by Search in all subsequent tool calls for this task. If omitted, a new session is created.
  • search — Discover AIsa tools and Plans for external data or actions. Skip tasks that only need supplied content or general knowledge. For multi-step tasks, search the full goal before decomposing it. Use English; put identifiers in known_fields. If has_full_schema=false, call batch_get_schema. Next call batch_quote, not batch_use. Use an identifier the user supplied, such as a ticker, exactly as given, even when it does not look like a conventional value; do not replace it with a more familiar identifier. Do not search for standalone rewriting, translation, summarization, formatting, or calculations on supplied content. For one capability, use an atomic query. Select a matching capability, date window, and metric. A Plan may cover only part of the task; assess its purpose and follow next_steps_guidance. Reuse the session_id returned by Search in all subsequent tool calls for this task. If omitted, a new session is created.

Data as of 2026-10-12T05:06:02+00:00.

OpenApps records store listings and observed changes. Store ranks are not download counts; launch dates may be derived or use the first observation. Coverage gaps are shown in the interactive catalog.

REST API specification · Guide for AI clients · Manage API keys and MCP connections