1package tools
2
3import (
4 "context"
5 "encoding/json"
6 "fmt"
7 "io"
8 "net/http"
9 "strings"
10 "time"
11
12 md "github.com/JohannesKaufmann/html-to-markdown"
13 "github.com/PuerkitoBio/goquery"
14 "github.com/kujtimiihoxha/termai/internal/config"
15 "github.com/kujtimiihoxha/termai/internal/permission"
16)
17
18type FetchParams struct {
19 URL string `json:"url"`
20 Format string `json:"format"`
21 Timeout int `json:"timeout,omitempty"`
22}
23
24type FetchPermissionsParams struct {
25 URL string `json:"url"`
26 Format string `json:"format"`
27 Timeout int `json:"timeout,omitempty"`
28}
29
30type fetchTool struct {
31 client *http.Client
32 permissions permission.Service
33}
34
35const (
36 FetchToolName = "fetch"
37 fetchToolDescription = `Fetches content from a URL and returns it in the specified format.
38
39WHEN TO USE THIS TOOL:
40- Use when you need to download content from a URL
41- Helpful for retrieving documentation, API responses, or web content
42- Useful for getting external information to assist with tasks
43
44HOW TO USE:
45- Provide the URL to fetch content from
46- Specify the desired output format (text, markdown, or html)
47- Optionally set a timeout for the request
48
49FEATURES:
50- Supports three output formats: text, markdown, and html
51- Automatically handles HTTP redirects
52- Sets reasonable timeouts to prevent hanging
53- Validates input parameters before making requests
54
55LIMITATIONS:
56- Maximum response size is 5MB
57- Only supports HTTP and HTTPS protocols
58- Cannot handle authentication or cookies
59- Some websites may block automated requests
60
61TIPS:
62- Use text format for plain text content or simple API responses
63- Use markdown format for content that should be rendered with formatting
64- Use html format when you need the raw HTML structure
65- Set appropriate timeouts for potentially slow websites`
66)
67
68func NewFetchTool(permissions permission.Service) BaseTool {
69 return &fetchTool{
70 client: &http.Client{
71 Timeout: 30 * time.Second,
72 },
73 permissions: permissions,
74 }
75}
76
77func (t *fetchTool) Info() ToolInfo {
78 return ToolInfo{
79 Name: FetchToolName,
80 Description: fetchToolDescription,
81 Parameters: map[string]any{
82 "url": map[string]any{
83 "type": "string",
84 "description": "The URL to fetch content from",
85 },
86 "format": map[string]any{
87 "type": "string",
88 "description": "The format to return the content in (text, markdown, or html)",
89 "enum": []string{"text", "markdown", "html"},
90 },
91 "timeout": map[string]any{
92 "type": "number",
93 "description": "Optional timeout in seconds (max 120)",
94 },
95 },
96 Required: []string{"url", "format"},
97 }
98}
99
100func (t *fetchTool) Run(ctx context.Context, call ToolCall) (ToolResponse, error) {
101 var params FetchParams
102 if err := json.Unmarshal([]byte(call.Input), ¶ms); err != nil {
103 return NewTextErrorResponse("Failed to parse fetch parameters: " + err.Error()), nil
104 }
105
106 if params.URL == "" {
107 return NewTextErrorResponse("URL parameter is required"), nil
108 }
109
110 format := strings.ToLower(params.Format)
111 if format != "text" && format != "markdown" && format != "html" {
112 return NewTextErrorResponse("Format must be one of: text, markdown, html"), nil
113 }
114
115 if !strings.HasPrefix(params.URL, "http://") && !strings.HasPrefix(params.URL, "https://") {
116 return NewTextErrorResponse("URL must start with http:// or https://"), nil
117 }
118
119 p := t.permissions.Request(
120 permission.CreatePermissionRequest{
121 Path: config.WorkingDirectory(),
122 ToolName: FetchToolName,
123 Action: "fetch",
124 Description: fmt.Sprintf("Fetch content from URL: %s", params.URL),
125 Params: FetchPermissionsParams(params),
126 },
127 )
128
129 if !p {
130 return ToolResponse{}, permission.ErrorPermissionDenied
131 }
132
133 client := t.client
134 if params.Timeout > 0 {
135 maxTimeout := 120 // 2 minutes
136 if params.Timeout > maxTimeout {
137 params.Timeout = maxTimeout
138 }
139 client = &http.Client{
140 Timeout: time.Duration(params.Timeout) * time.Second,
141 }
142 }
143
144 req, err := http.NewRequestWithContext(ctx, "GET", params.URL, nil)
145 if err != nil {
146 return ToolResponse{}, fmt.Errorf("failed to create request: %w", err)
147 }
148
149 req.Header.Set("User-Agent", "termai/1.0")
150
151 resp, err := client.Do(req)
152 if err != nil {
153 return ToolResponse{}, fmt.Errorf("failed to fetch URL: %w", err)
154 }
155 defer resp.Body.Close()
156
157 if resp.StatusCode != http.StatusOK {
158 return NewTextErrorResponse(fmt.Sprintf("Request failed with status code: %d", resp.StatusCode)), nil
159 }
160
161 maxSize := int64(5 * 1024 * 1024) // 5MB
162 body, err := io.ReadAll(io.LimitReader(resp.Body, maxSize))
163 if err != nil {
164 return NewTextErrorResponse("Failed to read response body: " + err.Error()), nil
165 }
166
167 content := string(body)
168 contentType := resp.Header.Get("Content-Type")
169
170 switch format {
171 case "text":
172 if strings.Contains(contentType, "text/html") {
173 text, err := extractTextFromHTML(content)
174 if err != nil {
175 return NewTextErrorResponse("Failed to extract text from HTML: " + err.Error()), nil
176 }
177 return NewTextResponse(text), nil
178 }
179 return NewTextResponse(content), nil
180
181 case "markdown":
182 if strings.Contains(contentType, "text/html") {
183 markdown, err := convertHTMLToMarkdown(content)
184 if err != nil {
185 return NewTextErrorResponse("Failed to convert HTML to Markdown: " + err.Error()), nil
186 }
187 return NewTextResponse(markdown), nil
188 }
189
190 return NewTextResponse("```\n" + content + "\n```"), nil
191
192 case "html":
193 return NewTextResponse(content), nil
194
195 default:
196 return NewTextResponse(content), nil
197 }
198}
199
200func extractTextFromHTML(html string) (string, error) {
201 doc, err := goquery.NewDocumentFromReader(strings.NewReader(html))
202 if err != nil {
203 return "", err
204 }
205
206 text := doc.Text()
207 text = strings.Join(strings.Fields(text), " ")
208
209 return text, nil
210}
211
212func convertHTMLToMarkdown(html string) (string, error) {
213 converter := md.NewConverter("", true, nil)
214
215 markdown, err := converter.ConvertString(html)
216 if err != nil {
217 return "", err
218 }
219
220 return markdown, nil
221}