@ctxprotocol/sdk 0.13.1 → 0.15.0

package/README.md

@@ -1,32 +1,32 @@
 # @ctxprotocol/sdk
 
-**The Universal Adapter for AI Agents.**
+**Grounded context for your AI agents — on demand.**
 
-Connect your AI to the real world without managing API keys, hosting servers, or reading documentation.
+Give your agent the live facts it's missing. Context is a pay-per-response data marketplace: your agent discovers tools at runtime, buys the data its question needs, and drops it straight into its context window — no API keys, no servers, no per-source integration work.
 
-Context Protocol is **npm for AI capabilities**. Just as you install packages to add functionality to your code, use the Context SDK to give your Agent instant access to thousands of live data sources and actions—from DeFi and Gas Oracles to Weather and Search.
+Context Protocol is **npm for AI capabilities**. Just as you install packages to add functionality to your code, use the Context SDK to give your Agent instant access to premium data feeds and live tools: the market intelligence usually locked behind steep monthly or annual subscriptions, unbundled into pay-per-response, from crypto and on-chain analytics to gas oracles, weather, and search.
 
 [![npm version](https://img.shields.io/npm/v/@ctxprotocol/sdk.svg)](https://www.npmjs.com/package/@ctxprotocol/sdk)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
 ---
 
-### 💰 $10,000 Developer Grant Program
+### Contributors
 
-We're funding the initial supply of MCP Tools for the Context Marketplace. **Become a Data Broker.**
+List MCP tools on the permissionless Context marketplace and keep **90% of usage fees** in USDC.
 
-- **🛠️ Build:** Create an MCP Server using this SDK (Solana data, Trading tools, Scrapers, etc.)
-- **📦 List:** Publish it to the Context Registry
-- **💵 Earn:** Get a **$250–$1,000 Grant** + earn USDC every time an agent queries your tool
+- **Build:** Create an MCP server with this SDK
+- **List:** Register at [ctxprotocol.com/contribute](https://ctxprotocol.com/contribute)
+- **Earn:** Revenue share on every Query response and eligible Execute call
 
-👉 [**View Open Bounties & Apply Here**](https://docs.ctxprotocol.com/grants)
+Grant batches open periodically — see [Developer Grants](https://docs.ctxprotocol.com/grants) for current batch status and contributor guidance.
 
 ---
 
 ## Why use Context?
 
 - **🔌 One Interface, Everything:** Stop integrating APIs one by one. Use a single SDK to access any tool in the marketplace.
-- **🧠 Zero-Ops:** We're a gateway to the best MCP tools. Just send the JSON and get the result.
+- **🧠 Zero-Ops:** We're the gateway to premium data feeds usually locked behind steep subscriptions. We host the connections; you just send the JSON and get the result.
 - **⚡️ Agentic Discovery:** Your Agent can search the marketplace at runtime to find tools it didn't know it needed.
 - **💸 Dual-Surface Economics:** Use Query for pay-per-response intelligence or Execute for session-budgeted method calls.
 
@@ -96,7 +96,7 @@ const result = await client.tools.execute({
 console.log(result.session); // methodPrice, spent, remaining, maxSpend, ...
 ```
 
-**Query mode** gives you a managed librarian contract — the server runs the live pipeline (`discover -> select -> metadata scout -> clarify if needed -> iterative execute -> synthesize -> settle`) with model-aware context budgeting and can return plain answers or structured evidence packages for one flat fee:
+**Query mode** gives you a managed librarian contract — the server runs the live pipeline (`discover -> select -> metadata scout -> iterative execute -> synthesize -> settle`) with model-aware context budgeting and can return plain answers or structured evidence packages for one flat fee:
 ```typescript
 const answer = await client.query.run({
   query: "What are the top whale movements on Base?",
@@ -417,7 +417,7 @@ const closed = await client.tools.closeSession("sess_123");
 
 #### `client.query.run(options)`
 
-Run an agentic query. The server applies the live librarian pipeline (`discover -> select -> metadata scout -> clarify if needed -> iterative execute -> synthesize -> settle`) with up to 100 MCP calls per response turn, then returns the selected Query response contract (`answer`, `answer_with_evidence`, or `evidence_only`).
+Run an agentic query. The server applies the live librarian pipeline (`discover -> select -> metadata scout -> iterative execute -> synthesize -> settle`) with up to 100 MCP calls per response turn, then returns the selected Query response contract (`answer`, `answer_with_evidence`, or `evidence_only`).
 
 The query runtime now exposes a single managed executor surface.
 The server decides internal budgets, ambiguity handling, and exploration policy

package/dist/client/index.cjs

@@ -53,9 +53,9 @@ var Developer = class {
     if (!toolId) {
       throw new ContextError("toolId is required");
     }
-    if (updates.name === void 0 && updates.description === void 0 && updates.category === void 0) {
+    if (updates.name === void 0 && updates.description === void 0 && updates.suggestedPrompts === void 0 && updates.category === void 0) {
       throw new ContextError(
-        "At least one field required: name, description, or category"
+        "At least one field required: name, description, suggestedPrompts, or category"
       );
     }
     if (updates.category !== void 0 && updates.category !== null && !ALLOWED_TOOL_CATEGORIES.includes(updates.category)) {
@@ -304,9 +304,6 @@ var Query = class {
   }
   normalizeResult(result) {
     const candidate = result;
-    if (candidate.outcomeType === "clarification_required" && "clarification" in candidate && candidate.clarification) {
-      return candidate;
-    }
     if (candidate.outcomeType === "capability_miss" && "capabilityMiss" in candidate && candidate.capabilityMiss) {
       return candidate;
     }
@@ -315,34 +312,6 @@ var Query = class {
       outcomeType: "answer"
     };
   }
-  buildPolicyErrorEvent(params) {
-    if (params.clarificationPolicy !== "error") {
-      return;
-    }
-    if (params.result.outcomeType === "clarification_required") {
-      return {
-        type: "error",
-        error: params.result.response,
-        code: "clarification_required",
-        reasonCode: "clarification_required",
-        outcomeType: "clarification_required",
-        clarification: params.result.clarification,
-        querySession: params.result.querySession
-      };
-    }
-    if (params.result.outcomeType === "capability_miss") {
-      return {
-        type: "error",
-        error: params.result.response,
-        code: "capability_miss",
-        reasonCode: "capability_miss",
-        outcomeType: "capability_miss",
-        capabilityMiss: params.result.capabilityMiss,
-        querySession: params.result.querySession
-      };
-    }
-    return void 0;
-  }
   buildSyntheticTraceFromRunResult(params) {
     const timeline = params.toolsUsed.map((tool, index) => ({
       stepType: "tool-call",
@@ -556,7 +525,6 @@ var Query = class {
         tools: opts.tools,
         resumeFrom: opts.resumeFrom,
         forkFrom: opts.forkFrom,
-        clarificationPolicy: opts.clarificationPolicy,
         answerModelId: opts.answerModelId,
         responseShape: opts.responseShape,
         favoritesOnly: opts.favoritesOnly,
@@ -614,13 +582,6 @@ var Query = class {
           normalizedResult.developerTrace = mergedTrace;
         }
         event.result = normalizedResult;
-        const policyErrorEvent = this.buildPolicyErrorEvent({
-          result: normalizedResult,
-          clarificationPolicy: opts.clarificationPolicy
-        });
-        if (policyErrorEvent) {
-          return policyErrorEvent;
-        }
       }
       return event;
     };
@@ -691,7 +652,7 @@ var ContextClient = class {
    *
    * Unlike `tools.execute()` which calls a single tool once, `query` sends
    * a natural-language question and lets the server handle discovery,
-   * metadata scout, clarification, iterative execution, and AI synthesis —
+   * metadata scout, iterative execution, and AI synthesis —
    * one flat fee.
    */
   query;