page-agent 1.5.2 → 1.5.3

package/README.md

@@ -11,7 +11,7 @@ The GUI Agent Living in Your Webpage. Control web interfaces with natural langua
 
 🌐 **English** | [中文](./docs/README-zh.md)
 
-👉 <a href="https://alibaba.github.io/page-agent/" target="_blank"><b>🚀 Demo</b></a> | <a href="https://alibaba.github.io/page-agent/docs/introduction/overview" target="_blank"><b>📖 Documentation</b></a>
+👉 <a href="https://alibaba.github.io/page-agent/" target="_blank"><b>🚀 Demo</b></a> | <a href="https://alibaba.github.io/page-agent/docs/introduction/overview" target="_blank"><b>📖 Documentation</b></a> | <a href="https://news.ycombinator.com/item?id=47264138" target="_blank">📢 Join HN Discussion</a>
 
 <video id="demo-video" src="https://github.com/user-attachments/assets/a1f2eae2-13fb-4aae-98cf-a3fc1620a6c2" controls crossorigin muted></video>
 
@@ -49,8 +49,8 @@ Fastest way to try PageAgent with our free Demo LLM:
 
 | Mirrors | URL                                                                                |
 | ------- | ---------------------------------------------------------------------------------- |
-| Global  | https://cdn.jsdelivr.net/npm/page-agent@1.5.2/dist/iife/page-agent.demo.js         |
-| China   | https://registry.npmmirror.com/page-agent/1.5.2/files/dist/iife/page-agent.demo.js |
+| Global  | https://cdn.jsdelivr.net/npm/page-agent@1.5.3/dist/iife/page-agent.demo.js         |
+| China   | https://registry.npmmirror.com/page-agent/1.5.3/files/dist/iife/page-agent.demo.js |
 
 > **⚠️ For technical evaluation only.** This demo CDN uses our free [testing LLM API](https://alibaba.github.io/page-agent/docs/features/models#free-testing-api). By using it, you agree to its [terms](https://github.com/alibaba/page-agent/blob/main/docs/terms-and-privacy.md).
 
@@ -111,3 +111,11 @@ file and in the node_modules directory after installation.
 ---
 
 **⭐ Star this repo if you find PageAgent helpful!**
+
+<a href="https://www.star-history.com/?repos=alibaba%2Fpage-agent&type=date">
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/image?repos=alibaba/page-agent&type=date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/image?repos=alibaba/page-agent&type=date" />
+   <img alt="Star History Chart" src="https://api.star-history.com/image?repos=alibaba/page-agent&type=date" />
+ </picture>
+</a>

package/dist/iife/page-agent.demo.js

@@ -42,7 +42,7 @@ Set the \`cycles\` parameter to \`"ref"\` to resolve cyclical schemas with defs.
 `:`
 `)+o,r=n+1,n=e.indexOf(`
 `,r)}while(n!==-1);return i+=e.slice(r),i}a(stringEncaseCRLFWithFirstIndex,"stringEncaseCRLFWithFirstIndex");const{stdout:stdoutColor,stderr:stderrColor}=supportsColor,GENERATOR=Symbol("GENERATOR"),STYLER=Symbol("STYLER"),IS_EMPTY=Symbol("IS_EMPTY"),levelMapping=["ansi","ansi","ansi256","ansi16m"],styles$2=Object.create(null),applyOptions=a((e,t={})=>{if(t.level&&!(Number.isInteger(t.level)&&t.level>=0&&t.level<=3))throw new Error("The `level` option should be an integer from 0 to 3");const o=stdoutColor?stdoutColor.level:0;e.level=t.level===void 0?o:t.level},"applyOptions"),chalkFactory=a(e=>{const t=a((...o)=>o.join(" "),"chalk");return applyOptions(t,e),Object.setPrototypeOf(t,createChalk.prototype),t},"chalkFactory");function createChalk(e){return chalkFactory(e)}a(createChalk,"createChalk"),Object.setPrototypeOf(createChalk.prototype,Function.prototype);for(const[e,t]of Object.entries(ansiStyles))styles$2[e]={get(){const o=createBuilder(this,createStyler(t.open,t.close,this[STYLER]),this[IS_EMPTY]);return Object.defineProperty(this,e,{value:o}),o}};styles$2.visible={get(){const e=createBuilder(this,this[STYLER],!0);return Object.defineProperty(this,"visible",{value:e}),e}};const getModelAnsi=a((e,t,o,...n)=>e==="rgb"?t==="ansi16m"?ansiStyles[o].ansi16m(...n):t==="ansi256"?ansiStyles[o].ansi256(ansiStyles.rgbToAnsi256(...n)):ansiStyles[o].ansi(ansiStyles.rgbToAnsi(...n)):e==="hex"?getModelAnsi("rgb",t,o,...ansiStyles.hexToRgb(...n)):ansiStyles[o][e](...n),"getModelAnsi"),usedModels=["rgb","hex","ansi256"];for(const e of usedModels){styles$2[e]={get(){const{level:o}=this;return function(...n){const r=createStyler(getModelAnsi(e,levelMapping[o],"color",...n),ansiStyles.color.close,this[STYLER]);return createBuilder(this,r,this[IS_EMPTY])}}};const t="bg"+e[0].toUpperCase()+e.slice(1);styles$2[t]={get(){const{level:o}=this;return function(...n){const r=createStyler(getModelAnsi(e,levelMapping[o],"bgColor",...n),ansiStyles.bgColor.close,this[STYLER]);return createBuilder(this,r,this[IS_EMPTY])}}}}const proto=Object.defineProperties(()=>{},{...styles$2,level:{enumerable:!0,get(){return this[GENERATOR].level},set(e){this[GENERATOR].level=e}}}),createStyler=a((e,t,o)=>{let n,r;return o===void 0?(n=e,r=t):(n=o.openAll+e,r=t+o.closeAll),{open:e,close:t,openAll:n,closeAll:r,parent:o}},"createStyler"),createBuilder=a((e,t,o)=>{const n=a((...r)=>applyStyle(n,r.length===1?""+r[0]:r.join(" ")),"builder");return Object.setPrototypeOf(n,proto),n[GENERATOR]=e,n[STYLER]=t,n[IS_EMPTY]=o,n},"createBuilder"),applyStyle=a((e,t)=>{if(e.level<=0||!t)return e[IS_EMPTY]?"":t;let o=e[STYLER];if(o===void 0)return t;const{openAll:n,closeAll:r}=o;if(t.includes("\x1B"))for(;o!==void 0;)t=stringReplaceAll(t,o.close,o.open),o=o.parent;const i=t.indexOf(`
-`);return i!==-1&&(t=stringEncaseCRLFWithFirstIndex(t,r,n,i)),n+t+r},"applyStyle");Object.defineProperties(createChalk.prototype,styles$2);const chalk=createChalk();createChalk({level:stderrColor?stderrColor.level:0});const debug=console.debug.bind(console,chalk.gray("[LLM]"));function zodToOpenAITool(e,t){return{type:"function",function:{name:e,description:t.description,parameters:toJSONSchema(t.inputSchema,{target:"openapi-3.0"})}}}a(zodToOpenAITool,"zodToOpenAITool");function modelPatch(e){const t=e.model||"";if(!t)return e;const o=normalizeModelName(t);return o.startsWith("qwen")&&(debug("Applying Qwen patch: use higher temperature for auto fixing"),e.temperature=Math.max(e.temperature||0,1),e.enable_thinking=!1),o.startsWith("claude")&&(debug("Applying Claude patch: disable thinking"),e.thinking={type:"disabled"},e.tool_choice==="required"?(debug('Applying Claude patch: convert tool_choice "required" to { type: "any" }'),e.tool_choice={type:"any"}):e.tool_choice?.function?.name&&(debug("Applying Claude patch: convert tool_choice format"),e.tool_choice={type:"tool",name:e.tool_choice.function.name})),o.startsWith("grok")&&(debug("Applying Grok patch: removing tool_choice"),delete e.tool_choice,debug("Applying Grok patch: disable reasoning and thinking"),e.thinking={type:"disabled",effort:"minimal"},e.reasoning={enabled:!1,effort:"low"}),o.startsWith("gpt")&&(debug("Applying GPT patch: set verbosity to low"),e.verbosity="low",o.startsWith("gpt-52")?(debug("Applying GPT-52 patch: disable reasoning"),e.reasoning_effort="none"):o.startsWith("gpt-51")?(debug("Applying GPT-51 patch: disable reasoning"),e.reasoning_effort="none"):o.startsWith("gpt-5-mini")?(debug("Applying GPT-5-mini patch: set reasoning effort to low, temperature to 1"),e.reasoning_effort="low",e.temperature=1):o.startsWith("gpt-5")&&(debug("Applying GPT-5 patch: set reasoning effort to low"),e.reasoning_effort="low")),o.startsWith("gemini")&&(debug("Applying Gemini patch: set reasoning effort to minimal"),e.reasoning_effort="minimal"),e}a(modelPatch,"modelPatch");function normalizeModelName(e){let t=e.toLowerCase();return t.includes("/")&&(t=t.split("/")[1]),t=t.replace(/_/g,""),t=t.replace(/\./g,""),t}a(normalizeModelName,"normalizeModelName");const ct=class ct{config;fetch;constructor(t){this.config=t,this.fetch=t.customFetch}async invoke(t,o,n,r){const i=Object.entries(o).map(([k,v])=>zodToOpenAITool(k,v)),s={model:this.config.model,temperature:this.config.temperature,messages:t,tools:i,parallel_tool_calls:!1,tool_choice:r?.toolChoiceName?{type:"function",function:{name:r.toolChoiceName}}:"required"};modelPatch(s);let c;try{c=await this.fetch(`${this.config.baseURL}/chat/completions`,{method:"POST",headers:{"Content-Type":"application/json",Authorization:`Bearer ${this.config.apiKey}`},body:JSON.stringify(s),signal:n})}catch(k){const v=k?.name==="AbortError",O=v?"Network request aborted":"Network request failed";throw v||console.error(k),new InvokeError(InvokeErrorType.NETWORK_ERROR,O,k)}if(!c.ok){const k=await c.json().catch(),v=k.error?.message||c.statusText;throw c.status===401||c.status===403?new InvokeError(InvokeErrorType.AUTH_ERROR,`Authentication failed: ${v}`,k):c.status===429?new InvokeError(InvokeErrorType.RATE_LIMIT,`Rate limit exceeded: ${v}`,k):c.status>=500?new InvokeError(InvokeErrorType.SERVER_ERROR,`Server error: ${v}`,k):new InvokeError(InvokeErrorType.UNKNOWN,`HTTP ${c.status}: ${v}`,k)}const l=await c.json(),d=l.choices?.[0];if(!d)throw new InvokeError(InvokeErrorType.UNKNOWN,"No choices in response",l);switch(d.finish_reason){case"tool_calls":case"function_call":case"stop":break;case"length":throw new InvokeError(InvokeErrorType.CONTEXT_LENGTH,"Response truncated: max tokens reached",void 0,l);case"content_filter":throw new InvokeError(InvokeErrorType.CONTENT_FILTER,"Content filtered by safety system",void 0,l);default:throw new InvokeError(InvokeErrorType.UNKNOWN,`Unexpected finish_reason: ${d.finish_reason}`,void 0,l)}const f=(r?.normalizeResponse?r.normalizeResponse(l):l).choices?.[0],g=f?.message?.tool_calls?.[0]?.function?.name;if(!g)throw new InvokeError(InvokeErrorType.NO_TOOL_CALL,"No tool call found in response",void 0,l);const m=o[g];if(!m)throw new InvokeError(InvokeErrorType.UNKNOWN,`Tool "${g}" not found in tools`,void 0,l);const T=f.message?.tool_calls?.[0]?.function?.arguments;if(!T)throw new InvokeError(InvokeErrorType.INVALID_TOOL_ARGS,"No tool call arguments found",void 0,l);let I;try{I=JSON.parse(T)}catch(k){throw new InvokeError(InvokeErrorType.INVALID_TOOL_ARGS,"Failed to parse tool arguments as JSON",k,l)}const A=m.inputSchema.safeParse(I);if(!A.success)throw console.error(prettifyError(A.error)),new InvokeError(InvokeErrorType.INVALID_TOOL_ARGS,"Tool arguments validation failed",A.error,l);const C=A.data;let D;try{D=await m.execute(C)}catch(k){throw new InvokeError(InvokeErrorType.TOOL_EXECUTION_ERROR,`Tool execution failed: ${k.message}`,k,l)}return{toolCall:{name:g,args:C},toolResult:D,usage:{promptTokens:l.usage?.prompt_tokens??0,completionTokens:l.usage?.completion_tokens??0,totalTokens:l.usage?.total_tokens??0,cachedTokens:l.usage?.prompt_tokens_details?.cached_tokens,reasoningTokens:l.usage?.completion_tokens_details?.reasoning_tokens},rawResponse:l,rawRequest:s}}};a(ct,"OpenAIClient");let OpenAIClient=ct;const LLM_MAX_RETRIES=2,DEFAULT_TEMPERATURE=.7;function parseLLMConfig(e){if(!e.baseURL||!e.apiKey||!e.model)throw new Error("[PageAgent] LLM configuration required. Please provide: baseURL, apiKey, model. See: https://alibaba.github.io/page-agent/docs/features/models");return{baseURL:e.baseURL,apiKey:e.apiKey,model:e.model,temperature:e.temperature??DEFAULT_TEMPERATURE,maxRetries:e.maxRetries??LLM_MAX_RETRIES,customFetch:(e.customFetch??fetch).bind(globalThis)}}a(parseLLMConfig,"parseLLMConfig");const lt=class lt extends EventTarget{config;client;constructor(t){super(),this.config=parseLLMConfig(t),this.client=new OpenAIClient(this.config)}async invoke(t,o,n,r){return await withRetry(async()=>{if(n.aborted)throw new Error("AbortError");return await this.client.invoke(t,o,n,r)},{maxRetries:this.config.maxRetries,onRetry:a(i=>{this.dispatchEvent(new CustomEvent("retry",{detail:{attempt:i,maxAttempts:this.config.maxRetries}}))},"onRetry"),onError:a(i=>{this.dispatchEvent(new CustomEvent("error",{detail:{error:i}}))},"onError")})}};a(lt,"LLM");let LLM=lt;async function withRetry(e,t){let o=0,n=null;for(;o<=t.maxRetries;){o>0&&(t.onRetry(o),await new Promise(r=>setTimeout(r,100)));try{return await e()}catch(r){if(r?.rawError?.name==="AbortError"||(console.error(r),t.onError(r),r instanceof InvokeError&&!r.retryable))throw r;n=r,o++,await new Promise(i=>setTimeout(i,100))}}throw n}a(withRetry,"withRetry");const SYSTEM_PROMPT=`You are an AI agent designed to operate in an iterative loop to automate browser tasks. Your ultimate goal is accomplishing the task provided in <user_request>.
+`);return i!==-1&&(t=stringEncaseCRLFWithFirstIndex(t,r,n,i)),n+t+r},"applyStyle");Object.defineProperties(createChalk.prototype,styles$2);const chalk=createChalk();createChalk({level:stderrColor?stderrColor.level:0});const debug=console.debug.bind(console,chalk.gray("[LLM]"));function zodToOpenAITool(e,t){return{type:"function",function:{name:e,description:t.description,parameters:toJSONSchema(t.inputSchema,{target:"openapi-3.0"})}}}a(zodToOpenAITool,"zodToOpenAITool");function modelPatch(e){const t=e.model||"";if(!t)return e;const o=normalizeModelName(t);return o.startsWith("qwen")&&(debug("Applying Qwen patch: use higher temperature for auto fixing"),e.temperature=Math.max(e.temperature||0,1),e.enable_thinking=!1),o.startsWith("claude")&&(debug("Applying Claude patch: disable thinking"),e.thinking={type:"disabled"},e.tool_choice==="required"?(debug('Applying Claude patch: convert tool_choice "required" to { type: "any" }'),e.tool_choice={type:"any"}):e.tool_choice?.function?.name&&(debug("Applying Claude patch: convert tool_choice format"),e.tool_choice={type:"tool",name:e.tool_choice.function.name})),o.startsWith("grok")&&(debug("Applying Grok patch: removing tool_choice"),delete e.tool_choice,debug("Applying Grok patch: disable reasoning and thinking"),e.thinking={type:"disabled",effort:"minimal"},e.reasoning={enabled:!1,effort:"low"}),o.startsWith("gpt")&&(debug("Applying GPT patch: set verbosity to low"),e.verbosity="low",o.startsWith("gpt-52")?(debug("Applying GPT-52 patch: disable reasoning"),e.reasoning_effort="none"):o.startsWith("gpt-51")?(debug("Applying GPT-51 patch: disable reasoning"),e.reasoning_effort="none"):o.startsWith("gpt-54")?(debug("Applying GPT-5.4 patch: skip reasoning_effort because chat/completions rejects it with function tools"),delete e.reasoning_effort):o.startsWith("gpt-5-mini")?(debug("Applying GPT-5-mini patch: set reasoning effort to low, temperature to 1"),e.reasoning_effort="low",e.temperature=1):o.startsWith("gpt-5")&&(debug("Applying GPT-5 patch: set reasoning effort to low"),e.reasoning_effort="low")),o.startsWith("gemini")&&(debug("Applying Gemini patch: set reasoning effort to minimal"),e.reasoning_effort="minimal"),e}a(modelPatch,"modelPatch");function normalizeModelName(e){let t=e.toLowerCase();return t.includes("/")&&(t=t.split("/")[1]),t=t.replace(/_/g,""),t=t.replace(/\./g,""),t}a(normalizeModelName,"normalizeModelName");const ct=class ct{config;fetch;constructor(t){this.config=t,this.fetch=t.customFetch}async invoke(t,o,n,r){const i=Object.entries(o).map(([k,v])=>zodToOpenAITool(k,v)),s={model:this.config.model,temperature:this.config.temperature,messages:t,tools:i,parallel_tool_calls:!1,tool_choice:r?.toolChoiceName?{type:"function",function:{name:r.toolChoiceName}}:"required"};modelPatch(s);let c;try{c=await this.fetch(`${this.config.baseURL}/chat/completions`,{method:"POST",headers:{"Content-Type":"application/json",Authorization:`Bearer ${this.config.apiKey}`},body:JSON.stringify(s),signal:n})}catch(k){const v=k?.name==="AbortError",O=v?"Network request aborted":"Network request failed";throw v||console.error(k),new InvokeError(InvokeErrorType.NETWORK_ERROR,O,k)}if(!c.ok){const k=await c.json().catch(),v=k.error?.message||c.statusText;throw c.status===401||c.status===403?new InvokeError(InvokeErrorType.AUTH_ERROR,`Authentication failed: ${v}`,k):c.status===429?new InvokeError(InvokeErrorType.RATE_LIMIT,`Rate limit exceeded: ${v}`,k):c.status>=500?new InvokeError(InvokeErrorType.SERVER_ERROR,`Server error: ${v}`,k):new InvokeError(InvokeErrorType.UNKNOWN,`HTTP ${c.status}: ${v}`,k)}const l=await c.json(),d=l.choices?.[0];if(!d)throw new InvokeError(InvokeErrorType.UNKNOWN,"No choices in response",l);switch(d.finish_reason){case"tool_calls":case"function_call":case"stop":break;case"length":throw new InvokeError(InvokeErrorType.CONTEXT_LENGTH,"Response truncated: max tokens reached",void 0,l);case"content_filter":throw new InvokeError(InvokeErrorType.CONTENT_FILTER,"Content filtered by safety system",void 0,l);default:throw new InvokeError(InvokeErrorType.UNKNOWN,`Unexpected finish_reason: ${d.finish_reason}`,void 0,l)}const f=(r?.normalizeResponse?r.normalizeResponse(l):l).choices?.[0],g=f?.message?.tool_calls?.[0]?.function?.name;if(!g)throw new InvokeError(InvokeErrorType.NO_TOOL_CALL,"No tool call found in response",void 0,l);const m=o[g];if(!m)throw new InvokeError(InvokeErrorType.UNKNOWN,`Tool "${g}" not found in tools`,void 0,l);const T=f.message?.tool_calls?.[0]?.function?.arguments;if(!T)throw new InvokeError(InvokeErrorType.INVALID_TOOL_ARGS,"No tool call arguments found",void 0,l);let I;try{I=JSON.parse(T)}catch(k){throw new InvokeError(InvokeErrorType.INVALID_TOOL_ARGS,"Failed to parse tool arguments as JSON",k,l)}const A=m.inputSchema.safeParse(I);if(!A.success)throw console.error(prettifyError(A.error)),new InvokeError(InvokeErrorType.INVALID_TOOL_ARGS,"Tool arguments validation failed",A.error,l);const C=A.data;let D;try{D=await m.execute(C)}catch(k){throw new InvokeError(InvokeErrorType.TOOL_EXECUTION_ERROR,`Tool execution failed: ${k.message}`,k,l)}return{toolCall:{name:g,args:C},toolResult:D,usage:{promptTokens:l.usage?.prompt_tokens??0,completionTokens:l.usage?.completion_tokens??0,totalTokens:l.usage?.total_tokens??0,cachedTokens:l.usage?.prompt_tokens_details?.cached_tokens,reasoningTokens:l.usage?.completion_tokens_details?.reasoning_tokens},rawResponse:l,rawRequest:s}}};a(ct,"OpenAIClient");let OpenAIClient=ct;const LLM_MAX_RETRIES=2,DEFAULT_TEMPERATURE=.7;function parseLLMConfig(e){if(!e.baseURL||!e.apiKey||!e.model)throw new Error("[PageAgent] LLM configuration required. Please provide: baseURL, apiKey, model. See: https://alibaba.github.io/page-agent/docs/features/models");return{baseURL:e.baseURL,apiKey:e.apiKey,model:e.model,temperature:e.temperature??DEFAULT_TEMPERATURE,maxRetries:e.maxRetries??LLM_MAX_RETRIES,customFetch:(e.customFetch??fetch).bind(globalThis)}}a(parseLLMConfig,"parseLLMConfig");const lt=class lt extends EventTarget{config;client;constructor(t){super(),this.config=parseLLMConfig(t),this.client=new OpenAIClient(this.config)}async invoke(t,o,n,r){return await withRetry(async()=>{if(n.aborted)throw new Error("AbortError");return await this.client.invoke(t,o,n,r)},{maxRetries:this.config.maxRetries,onRetry:a(i=>{this.dispatchEvent(new CustomEvent("retry",{detail:{attempt:i,maxAttempts:this.config.maxRetries}}))},"onRetry"),onError:a(i=>{this.dispatchEvent(new CustomEvent("error",{detail:{error:i}}))},"onError")})}};a(lt,"LLM");let LLM=lt;async function withRetry(e,t){let o=0,n=null;for(;o<=t.maxRetries;){o>0&&(t.onRetry(o),await new Promise(r=>setTimeout(r,100)));try{return await e()}catch(r){if(r?.rawError?.name==="AbortError"||(console.error(r),t.onError(r),r instanceof InvokeError&&!r.retryable))throw r;n=r,o++,await new Promise(i=>setTimeout(i,100))}}throw n}a(withRetry,"withRetry");const SYSTEM_PROMPT=`You are an AI agent designed to operate in an iterative loop to automate browser tasks. Your ultimate goal is accomplishing the task provided in <user_request>.
 
 <intro>
 You excel at following tasks:

package/package.json

@@ -1,7 +1,7 @@
 {
     "name": "page-agent",
     "private": false,
-    "version": "1.5.2",
+    "version": "1.5.3",
     "type": "module",
     "main": "./dist/esm/page-agent.js",
     "module": "./dist/esm/page-agent.js",
@@ -44,10 +44,10 @@
         "postpublish": "node -e \"['README.md','LICENSE'].forEach(f=>{try{require('fs').unlinkSync(f)}catch{}})\""
     },
     "dependencies": {
-        "@page-agent/core": "1.5.2",
-        "@page-agent/llms": "1.5.2",
-        "@page-agent/page-controller": "1.5.2",
-        "@page-agent/ui": "1.5.2",
+        "@page-agent/core": "1.5.3",
+        "@page-agent/llms": "1.5.3",
+        "@page-agent/page-controller": "1.5.3",
+        "@page-agent/ui": "1.5.3",
         "chalk": "^5.6.2"
     },
     "peerDependencies": {