Compute structured line-by-line text diff with optional unified format via @mukundakatta/diff-mcp

@mukundakatta/diff-mcp v0.1.1 — structured line-by-line text diff

Install: npm install @mukundakatta/diff-mcp Server: node node_modules/@mukundakatta/diff-mcp/dist/server.js (stdio)

Tools (1)

Output format

Returns {entries: [{op, text}, ...]} where op is "eq", "add", or "del". With unified: true, adds a unified string field with classic +/-/ prefix format.

Key gotchas

1. LINE DIFF ONLY — npm description claims "line/word/char" but the tool only supports line-by-line diff. There is NO mode parameter; passing one is silently ignored.
2. `unified: true` adds a rendered string alongside the structured entries — format is line (unchanged), -line (deleted), +line (added)
3. Whitespace changes ARE detected — "hello world" vs "hello world" shows as del+add (no whitespace normalization)
4. Identical strings return single {op: "eq"} entry
5. Empty→content shows del of empty string + add of new lines; content→empty shows del of lines + add of empty string
6. Trailing newline removal shows as del of empty string at end
7. Both empty returns {op: "eq", text: ""}
8. Unicode and emoji handled correctly — Turkish chars, emoji work fine

Verified trace (14 calls, 100% success, p50 < 1ms)

• diff({a: "hello\nworld\nfoo", b: "hello\nearth\nfoo"}) → [eq "hello", del "world", add "earth", eq "foo"]
• diff({a: "line1\nline2", b: "line1\ninserted\nline2\nappended"}) → [eq "line1", add "inserted", eq "line2", add "appended"]
• diff({a: "alpha\nbeta\ngamma\ndelta", b: "alpha\ndelta"}) → [eq "alpha", del "beta", del "gamma", eq "delta"]
• diff({a: "same text", b: "same text"}) → [eq "same text"]
• diff({a: "", b: "new content\nhere"}) → [del "", add "new content", add "here"]
• diff({a: "old content\nhere", b: ""}) → [del "old content", del "here", add ""]
• diff({a: "fn sub(a,b){return a-b}", b: "fn multiply(a,b){return a*b}"}) → correctly shows renamed function as del+add preserving unchanged surrounding
• diff({a: "hello world", b: "hello world"}) → [del "hello world", add "hello world"] (whitespace-sensitive)
• diff({a: "Merhaba dünya 🌍", b: "Merhaba dünya 🌎"}) → [del "Merhaba dünya 🌍", add "Merhaba dünya 🌎"] (line-level, not char)
• diff({a: "line1\nline2\nline3", b: "line1\nmodified\nline3\nline4", unified: true}) → entries + " line1\n-line2\n+modified\n line3\n+line4"