Skip to main content
Glama

edit_block

Destructive

Apply surgical edits to files by replacing specific text or updating Excel ranges with precise control over changes.

Instructions

                    Apply surgical edits to files.

                    BEST PRACTICE: Make multiple small, focused edits rather than one large edit.
                    Each edit_block call should change only what needs to be changed - include just enough
                    context to uniquely identify the text being modified.

                    FORMAT HANDLING (by extension):

                    EXCEL FILES (.xlsx, .xls, .xlsm) - Range Update mode:
                    Takes:
                    - file_path: Path to the Excel file
                    - range: ALWAYS use FROM:TO format - "SheetName!A1:C10" or "SheetName!C1:C1"
                    - content: 2D array, e.g., [["H1","H2"],["R1","R2"]]

                    TEXT FILES - Find/Replace mode:
                    Takes:
                    - file_path: Path to the file to edit
                    - old_string: Text to replace
                    - new_string: Replacement text
                    - expected_replacements: Optional number of replacements (default: 1)

                    By default, replaces only ONE occurrence of the search text.
                    To replace multiple occurrences, provide expected_replacements with
                    the exact number of matches expected.

                    UNIQUENESS REQUIREMENT: When expected_replacements=1 (default), include the minimal
                    amount of context necessary (typically 1-3 lines) before and after the change point,
                    with exact whitespace and indentation.

                    When editing multiple sections, make separate edit_block calls for each distinct change
                    rather than one large replacement.

                    When a close but non-exact match is found, a character-level diff is shown in the format:
                    common_prefix{-removed-}{+added+}common_suffix to help you identify what's different.

                    Similar to write_file, there is a configurable line limit (fileWriteLineLimit) that warns
                    if the edited file exceeds this limit. If this happens, consider breaking your edits into
                    smaller, more focused changes.

                    IMPORTANT: Always use absolute paths for reliability. Paths are automatically normalized regardless of slash direction. Relative paths may fail as they depend on the current working directory. Tilde paths (~/...) might not work in all contexts. Unless the user explicitly asks for relative paths, use absolute paths.
                    This command can be referenced as "DC: ..." or "use Desktop Commander to ..." in your instructions.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
file_pathYes
old_stringNo
new_stringNo
expected_replacementsNo
rangeNo
contentNo
optionsNo

Implementation Reference

  • The main handler function for the 'edit_block' tool. It parses arguments using EditBlockArgsSchema and dispatches to either structured file range editing (for Excel etc.) or text file search/replace editing.
    export async function handleEditBlock(args: unknown): Promise<ServerResult> {
        const parsed = EditBlockArgsSchema.parse(args);
    
        // Structured files: Range rewrite
        if (parsed.range !== undefined && parsed.content !== undefined) {
            try {
                // Validate path before any filesystem operations
                const validatedPath = await validatePath(parsed.file_path);
    
                const { getFileHandler } = await import('../utils/files/factory.js');
                const handler = await getFileHandler(validatedPath);
    
                // Parse content if it's a JSON string (AI often sends arrays as JSON strings)
                let content = parsed.content;
                if (typeof content === 'string') {
                    try {
                        content = JSON.parse(content);
                    } catch {
                        // Leave as-is if not valid JSON - let handler decide
                    }
                }
    
                // Check if handler supports range editing
                if ('editRange' in handler && typeof handler.editRange === 'function') {
                    await handler.editRange(validatedPath, parsed.range, content, parsed.options);
                    return {
                        content: [{
                            type: "text",
                            text: `Successfully updated range ${parsed.range} in ${parsed.file_path}`
                        }],
                    };
                } else {
                    return {
                        content: [{
                            type: "text",
                            text: `Error: Range-based editing not supported for ${parsed.file_path}`
                        }],
                        isError: true
                    };
                }
            } catch (error) {
                const errorMessage = error instanceof Error ? error.message : String(error);
                return {
                    content: [{
                        type: "text",
                        text: `Error: ${errorMessage}`
                    }],
                    isError: true
                };
            }
        }
    
        // Text files: String replacement
        // Validate required parameters for text replacement
        if (parsed.old_string === undefined || parsed.new_string === undefined) {
            return {
                content: [{
                    type: "text",
                    text: `Error: Text replacement requires both old_string and new_string parameters`
                }],
                isError: true
            };
        }
    
        const searchReplace = {
            search: parsed.old_string,
            replace: parsed.new_string
        };
    
        return performSearchReplace(parsed.file_path, searchReplace, parsed.expected_replacements);
    }
  • Zod schema defining input parameters for edit_block tool, supporting both text replacement (old_string, new_string, expected_replacements) and structured range edits (range, content).
    export const EditBlockArgsSchema = z.object({
      file_path: z.string(),
      // Text file string replacement
      old_string: z.string().optional(),
      new_string: z.string().optional(),
      expected_replacements: z.number().optional().default(1),
      // Structured file range rewrite (Excel, etc.)
      range: z.string().optional(),
      content: z.any().optional(),
      options: z.record(z.any()).optional()
    }).refine(
      data => (data.old_string !== undefined && data.new_string !== undefined) ||
              (data.range !== undefined && data.content !== undefined),
      { message: "Must provide either (old_string + new_string) or (range + content)" }
    );
  • src/server.ts:674-724 (registration)
    Registration of 'edit_block' tool in the list_tools handler, defining name, description, input schema, and annotations.
        name: "edit_block",
        description: `
                Apply surgical edits to files.
    
                BEST PRACTICE: Make multiple small, focused edits rather than one large edit.
                Each edit_block call should change only what needs to be changed - include just enough
                context to uniquely identify the text being modified.
    
                FORMAT HANDLING (by extension):
    
                EXCEL FILES (.xlsx, .xls, .xlsm) - Range Update mode:
                Takes:
                - file_path: Path to the Excel file
                - range: ALWAYS use FROM:TO format - "SheetName!A1:C10" or "SheetName!C1:C1"
                - content: 2D array, e.g., [["H1","H2"],["R1","R2"]]
    
                TEXT FILES - Find/Replace mode:
                Takes:
                - file_path: Path to the file to edit
                - old_string: Text to replace
                - new_string: Replacement text
                - expected_replacements: Optional number of replacements (default: 1)
    
                By default, replaces only ONE occurrence of the search text.
                To replace multiple occurrences, provide expected_replacements with
                the exact number of matches expected.
    
                UNIQUENESS REQUIREMENT: When expected_replacements=1 (default), include the minimal
                amount of context necessary (typically 1-3 lines) before and after the change point,
                with exact whitespace and indentation.
    
                When editing multiple sections, make separate edit_block calls for each distinct change
                rather than one large replacement.
    
                When a close but non-exact match is found, a character-level diff is shown in the format:
                common_prefix{-removed-}{+added+}common_suffix to help you identify what's different.
    
                Similar to write_file, there is a configurable line limit (fileWriteLineLimit) that warns
                if the edited file exceeds this limit. If this happens, consider breaking your edits into
                smaller, more focused changes.
    
                ${PATH_GUIDANCE}
                ${CMD_PREFIX_DESCRIPTION}`,
        inputSchema: zodToJsonSchema(EditBlockArgsSchema),
        annotations: {
            title: "Edit Block",
            readOnlyHint: false,
            destructiveHint: true,
            openWorldHint: false,
        },
    },
  • Dispatch in CallToolRequestSchema handler that routes 'edit_block' calls to handlers.handleEditBlock.
    case "edit_block":
        result = await handlers.handleEditBlock(args);
        break;
  • Core helper function implementing text file editing logic: exact search/replace, fuzzy matching fallback, line ending handling, occurrence counting, and warnings for large edits.
    export async function performSearchReplace(filePath: string, block: SearchReplace, expectedReplacements: number = 1): Promise<ServerResult> {
        // Get file extension for telemetry using path module
        const fileExtension = path.extname(filePath).toLowerCase();
        
        // Capture file extension and string sizes in telemetry without capturing the file path
        capture('server_edit_block', {
            fileExtension: fileExtension,
            oldStringLength: block.search.length,
            oldStringLines: block.search.split('\n').length,
            newStringLength: block.replace.length,
            newStringLines: block.replace.split('\n').length,
            expectedReplacements: expectedReplacements
        });
        // Check for empty search string to prevent infinite loops
        if (block.search === "") {
        
            // Capture file extension in telemetry without capturing the file path
            capture('server_edit_block_empty_search', {fileExtension: fileExtension, expectedReplacements});
            return {
                content: [{ 
                    type: "text", 
                    text: "Empty search strings are not allowed. Please provide a non-empty string to search for."
                }],
            };
        }
        
    
        // Read file directly to preserve line endings - critical for edit operations
        const validPath = await validatePath(filePath);
        const content = await readFileInternal(validPath, 0, Number.MAX_SAFE_INTEGER);
        
        // Make sure content is a string
        if (typeof content !== 'string') {
            capture('server_edit_block_content_not_string', {fileExtension: fileExtension, expectedReplacements});
            throw new Error('Wrong content for file ' + filePath);
        }
        
        // Get the line limit from configuration
        const config = await configManager.getConfig();
        const MAX_LINES = config.fileWriteLineLimit ?? 50; // Default to 50 if not set
        
        // Detect file's line ending style
        const fileLineEnding = detectLineEnding(content);
        
        // Normalize search string to match file's line endings
        const normalizedSearch = normalizeLineEndings(block.search, fileLineEnding);
        
        // First try exact match
        let tempContent = content;
        let count = 0;
        let pos = tempContent.indexOf(normalizedSearch);
        
        while (pos !== -1) {
            count++;
            pos = tempContent.indexOf(normalizedSearch, pos + 1);
        }
        
        // If exact match found and count matches expected replacements, proceed with exact replacement
        if (count > 0 && count === expectedReplacements) {
            // Replace all occurrences
            let newContent = content;
            
            // If we're only replacing one occurrence, replace it directly
            if (expectedReplacements === 1) {
                const searchIndex = newContent.indexOf(normalizedSearch);
                newContent = 
                    newContent.substring(0, searchIndex) + 
                    normalizeLineEndings(block.replace, fileLineEnding) + 
                    newContent.substring(searchIndex + normalizedSearch.length);
            } else {
                // Replace all occurrences using split and join for multiple replacements
                newContent = newContent.split(normalizedSearch).join(normalizeLineEndings(block.replace, fileLineEnding));
            }
            
            // Check if search or replace text has too many lines
            const searchLines = block.search.split('\n').length;
            const replaceLines = block.replace.split('\n').length;
            const maxLines = Math.max(searchLines, replaceLines);
            let warningMessage = "";
            
            if (maxLines > MAX_LINES) {
                const problemText = searchLines > replaceLines ? 'search text' : 'replacement text';
                warningMessage = `\n\nWARNING: The ${problemText} has ${maxLines} lines (maximum: ${MAX_LINES}).
                
    RECOMMENDATION: For large search/replace operations, consider breaking them into smaller chunks with fewer lines.`;
            }
            
            await writeFile(filePath, newContent);
            capture('server_edit_block_exact_success', {fileExtension: fileExtension, expectedReplacements, hasWarning: warningMessage !== ""});
            return {
                content: [{ 
                    type: "text", 
                    text: `Successfully applied ${expectedReplacements} edit${expectedReplacements > 1 ? 's' : ''} to ${filePath}${warningMessage}` 
                }],
            };
        }
        
        // If exact match found but count doesn't match expected, inform the user
        if (count > 0 && count !== expectedReplacements) {
            capture('server_edit_block_unexpected_count', {fileExtension: fileExtension, expectedReplacements, expectedReplacementsCount: count});
            return {
                content: [{ 
                    type: "text", 
                    text: `Expected ${expectedReplacements} occurrences but found ${count} in ${filePath}. ` + 
                `Double check and make sure you understand all occurencies and if you want to replace all ${count} occurrences, set expected_replacements to ${count}. ` +
                `If there are many occurrancies and you want to change some of them and keep the rest. Do it one by one, by adding more lines around each occurrence.` +
    `If you want to replace a specific occurrence, make your search string more unique by adding more lines around search string.`
                }],
            };
        }
        
        // If exact match not found, try fuzzy search
        if (count === 0) {
            // Track fuzzy search time
            const startTime = performance.now();
            
            // Perform fuzzy search
            const fuzzyResult = recursiveFuzzyIndexOf(content, block.search);
            const similarity = getSimilarityRatio(block.search, fuzzyResult.value);
            
            // Calculate execution time in milliseconds
            const executionTime = performance.now() - startTime;
            
            // Generate diff and gather character code data
            const diff = highlightDifferences(block.search, fuzzyResult.value);
            
            // Count character codes in diff
            const characterCodeData = getCharacterCodeData(block.search, fuzzyResult.value);
            
            // Create comprehensive log entry
            const logEntry: FuzzySearchLogEntry = {
                timestamp: new Date(),
                searchText: block.search,
                foundText: fuzzyResult.value,
                similarity: similarity,
                executionTime: executionTime,
                exactMatchCount: count,
                expectedReplacements: expectedReplacements,
                fuzzyThreshold: FUZZY_THRESHOLD,
                belowThreshold: similarity < FUZZY_THRESHOLD,
                diff: diff,
                searchLength: block.search.length,
                foundLength: fuzzyResult.value.length,
                fileExtension: fileExtension,
                characterCodes: characterCodeData.report,
                uniqueCharacterCount: characterCodeData.uniqueCount,
                diffLength: characterCodeData.diffLength
            };
            
            // Log to file
            await fuzzySearchLogger.log(logEntry);
            
            // Combine all fuzzy search data for single capture
            const fuzzySearchData = {
                similarity: similarity,
                execution_time_ms: executionTime,
                search_length: block.search.length,
                file_size: content.length,
                threshold: FUZZY_THRESHOLD,
                found_text_length: fuzzyResult.value.length,
                character_codes: characterCodeData.report,
                unique_character_count: characterCodeData.uniqueCount,
                total_diff_length: characterCodeData.diffLength
            };
            
            // Check if the fuzzy match is "close enough"
            if (similarity >= FUZZY_THRESHOLD) {
                // Capture the fuzzy search event with all data
                capture('server_fuzzy_search_performed', fuzzySearchData);
                
                // If we allow fuzzy matches, we would make the replacement here
                // For now, we'll return a detailed message about the fuzzy match
                return {
                    content: [{ 
                        type: "text", 
                        text: `Exact match not found, but found a similar text with ${Math.round(similarity * 100)}% similarity (found in ${executionTime.toFixed(2)}ms):\n\n` +
                              `Differences:\n${diff}\n\n` +
                              `To replace this text, use the exact text found in the file.\n\n` +
                              `Log entry saved for analysis. Use the following command to check the log:\n` +
                              `Check log: ${await fuzzySearchLogger.getLogPath()}`
                    }],// TODO
                };
            } else {
                // If the fuzzy match isn't close enough
                // Still capture the fuzzy search event with all data
                capture('server_fuzzy_search_performed', {
                    ...fuzzySearchData,
                    below_threshold: true
                });
                
                return {
                    content: [{ 
                        type: "text", 
                        text: `Search content not found in ${filePath}. The closest match was "${fuzzyResult.value}" ` +
                              `with only ${Math.round(similarity * 100)}% similarity, which is below the ${Math.round(FUZZY_THRESHOLD * 100)}% threshold. ` +
                              `(Fuzzy search completed in ${executionTime.toFixed(2)}ms)\n\n` +
                              `Log entry saved for analysis. Use the following command to check the log:\n` +
                              `Check log: ${await fuzzySearchLogger.getLogPath()}`
                    }],
                };
            }
        }
        
        throw new Error("Unexpected error during search and replace operation.");
    }
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations indicate destructiveHint=true and readOnlyHint=false, which the description aligns with by describing file modifications. The description adds valuable behavioral context beyond annotations: it explains uniqueness requirements for text edits, character-level diff displays for non-exact matches, configurable line limits (referencing write_file), and path handling rules (absolute vs. relative). However, it doesn't detail error handling or specific rate limits, keeping it from a perfect score.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured with clear sections (BEST PRACTICE, FORMAT HANDLING, UNIQUENESS REQUIREMENT) and uses bullet points for readability. It is appropriately sized for a complex tool with 7 parameters and multiple modes. However, some redundancy exists (e.g., repeating path advice and edit focus), and the final paragraph about command references is slightly extraneous, preventing a perfect score.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (7 parameters, 0% schema coverage, no output schema, destructive operations), the description is highly complete. It covers all operational modes (Excel vs. text), parameter usage, best practices, error handling (diffs for non-exact matches), constraints (line limits, path requirements), and sibling tool context. No output schema exists, but the description adequately explains expected behaviors without needing to detail return values.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 0% schema description coverage, the description fully compensates by explaining all key parameters. It details file_path usage for both modes, old_string/new_string for text files, expected_replacements with defaults, and range/content for Excel files. It clarifies parameter dependencies (e.g., range and content for Excel, old_string/new_string for text) and provides examples (e.g., 'SheetName!A1:C10', 2D arrays), adding significant meaning beyond the bare schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: 'Apply surgical edits to files.' It specifies the verb ('apply surgical edits') and resource ('files'), distinguishing it from sibling tools like write_file (which creates/overwrites files) and read_file (which only reads). The description further elaborates on the surgical nature of edits, making the purpose highly specific and differentiated.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit guidance on when and how to use this tool versus alternatives. It states: 'When editing multiple sections, make separate edit_block calls for each distinct change rather than one large replacement,' and references write_file for context on line limits. It also distinguishes between Excel file edits (range update mode) and text file edits (find/replace mode), offering clear alternatives within the tool itself based on file type.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/wonderwhy-er/ClaudeComputerCommander'

If you have feedback or need assistance with the MCP directory API, please join our Discord server