Skip to content

🤖 Add image reading support to file_read tool #357

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

🤖 Add image reading support to file_read tool #357

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
dffabb0
🤖 Add image reading support to file_read tool
kylecarbs Oct 20, 2025
5dfc0e2
Fix ESLint: use optional chaining for mime type checks
kylecarbs Oct 20, 2025
0ca165b
Pass tools to convertToModelMessages to enable toModelOutput
kylecarbs Oct 20, 2025
c728067
Fix ESLint errors in convertMessages test
kylecarbs Oct 20, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
86 changes: 34 additions & 52 deletions bun.lock
View file
Open in desktop

Large diffs are not rendered by default.

71 changes: 26 additions & 45 deletions package.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,20 +1,9 @@
{
"name": "@coder/cmux",
"name": "cmux",
"version": "0.3.0",
"description": "cmux - coder multiplexer",
"author": "Coder",
"main": "dist/main.js",
"bin": {
"cmux": "dist/main.js"
},
"license": "AGPL-3.0-only",
"repository": {
"type": "git",
"url": "https://github.com/coder/cmux.git"
},
"publishConfig": {
"access": "public"
},
"scripts": {
"dev": "make dev",
"prebuild:main": "./scripts/generate-version.sh",
Expand Down Expand Up @@ -46,29 +35,44 @@
"dependencies": {
"@ai-sdk/anthropic": "^2.0.29",
"@ai-sdk/openai": "^2.0.52",
"@emotion/react": "^11.14.0",
"@emotion/styled": "^11.14.1",
"ai": "^5.0.72",
"ai-tokenizer": "^1.0.3",
"chalk": "^5.6.2",
"cors": "^2.8.5",
"cmdk": "^1.0.0",
"crc-32": "^1.2.2",
"diff": "^8.0.2",
"disposablestack": "^1.1.7",
"electron-updater": "^6.6.2",
"express": "^5.1.0",
"escape-html": "^1.0.3",
"jsonc-parser": "^3.3.1",
"lru-cache": "^11.2.2",
"markdown-it": "^14.1.0",
"mermaid": "^11.12.0",
"mime-types": "^3.0.1",
"minimist": "^1.2.8",
"posthog-js": "^1.276.0",
"react": "^18.2.0",
"react-compiler-runtime": "^1.0.0",
"react-dnd": "^16.0.1",
"react-dnd-html5-backend": "^16.0.1",
"react-dom": "^18.2.0",
"react-markdown": "^10.1.0",
"rehype-katex": "^7.0.1",
"rehype-raw": "^7.0.0",
"rehype-sanitize": "^6.0.0",
"remark-gfm": "^4.0.1",
"remark-math": "^6.0.0",
"shiki": "^3.13.0",
"source-map-support": "^0.5.21",
"undici": "^7.16.0",
"write-file-atomic": "^6.0.0",
"ws": "^8.18.3",
"zod": "^4.1.11",
"zod-to-json-schema": "^3.24.6"
},
"devDependencies": {
"@emotion/babel-plugin": "^11.13.5",
"@emotion/react": "^11.14.0",
"@emotion/styled": "^11.14.1",
"@eslint/js": "^9.36.0",
"@playwright/test": "^1.56.0",
"@storybook/addon-essentials": "^8.6.14",
Expand All @@ -87,6 +91,7 @@
"@types/jest": "^30.0.0",
"@types/katex": "^0.16.7",
"@types/markdown-it": "^14.1.2",
"@types/mime-types": "^3.0.1",
"@types/minimist": "^1.2.5",
"@types/react": "^18.2.0",
"@types/react-dom": "^18.2.0",
Expand All @@ -97,8 +102,8 @@
"@typescript/native-preview": "^7.0.0-dev.20251014.1",
"@vitejs/plugin-react": "^4.0.0",
"babel-plugin-react-compiler": "^1.0.0",
"cmdk": "^1.0.0",
"concurrently": "^8.2.0",
"cors": "^2.8.5",
"dotenv": "^17.2.3",
"electron": "^38.2.1",
"electron-builder": "^24.6.0",
Expand All @@ -107,44 +112,20 @@
"eslint": "^9.36.0",
"eslint-plugin-react": "^7.37.5",
"eslint-plugin-react-hooks": "^5.2.0",
"escape-html": "^1.0.3",
"express": "^5.1.0",
"jest": "^30.1.3",
"markdown-it": "^14.1.0",
"mermaid": "^11.12.0",
"playwright": "^1.56.0",
"posthog-js": "^1.276.0",
"prettier": "^3.6.2",
"react": "^18.2.0",
"react-compiler-runtime": "^1.0.0",
"react-dnd": "^16.0.1",
"react-dnd-html5-backend": "^16.0.1",
"react-dom": "^18.2.0",
"react-markdown": "^10.1.0",
"rehype-katex": "^7.0.1",
"rehype-raw": "^7.0.0",
"rehype-sanitize": "^6.0.0",
"remark-gfm": "^4.0.1",
"remark-math": "^6.0.0",
"shiki": "^3.13.0",
"storybook": "^8.6.14",
"ts-jest": "^29.4.4",
"tsc-alias": "^1.8.16",
"typescript": "^5.1.3",
"typescript-eslint": "^8.45.0",
"vite": "^4.4.0",
"vite-plugin-svgr": "^4.5.0",
"vite-plugin-top-level-await": "^1.6.0"
"vite-plugin-top-level-await": "^1.6.0",
"ws": "^8.18.3"
},
"files": [
"dist/**/*.js",
"dist/**/*.js.map",
"dist/**/*.wasm",
"dist/**/*.html",
"dist/**/*.css",
"dist/assets/**/*",
"README.md",
"LICENSE"
],
"build": {
"appId": "com.cmux.app",
"productName": "cmux",
Expand Down
46 changes: 35 additions & 11 deletions src/components/tools/FileReadToolCall.tsx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -101,6 +101,14 @@ const InfoValue = styled.span`
word-break: break-all;
`;

const ImagePreview = styled.img`
max-width: 100%;
max-height: 400px;
border-radius: 3px;
display: block;
margin: 8px 0;
`;

interface FileReadToolCallProps {
args: FileReadToolArgs;
result?: FileReadToolResult;
Expand Down Expand Up @@ -170,7 +178,10 @@ export const FileReadToolCall: React.FC<FileReadToolCallProps> = ({
<Tooltip>file_read</Tooltip>
</TooltipWrapper>
<FilePathText>{filePath}</FilePathText>
{result && result.success && parsedContent && (
{result && result.success && result.mime_type?.startsWith("image/") && (
<MetadataText>{result.mime_type}</MetadataText>
)}
{result && result.success && parsedContent && !result.mime_type?.startsWith("image/") && (
<MetadataText>
read {formatBytes(parsedContent.actualBytes)} of {formatBytes(result.file_size)}
</MetadataText>
Expand Down Expand Up @@ -210,19 +221,32 @@ export const FileReadToolCall: React.FC<FileReadToolCallProps> = ({
</DetailSection>
)}

{result.success && result.content && parsedContent && (
{result.success && result.mime_type?.startsWith("image/") && (
<DetailSection>
<DetailLabel>Content</DetailLabel>
<ContentBlock>
<LineNumbers>
{parsedContent.lineNumbers.map((lineNum, i) => (
<div key={i}>{lineNum}</div>
))}
</LineNumbers>
<ContentText>{parsedContent.actualContent}</ContentText>
</ContentBlock>
<DetailLabel>Image Preview</DetailLabel>
<ImagePreview
src={`data:${result.mime_type};base64,${result.content}`}
alt={args.filePath}
/>
</DetailSection>
)}

{result.success &&
result.content &&
!result.mime_type?.startsWith("image/") &&
parsedContent && (
<DetailSection>
<DetailLabel>Content</DetailLabel>
<ContentBlock>
<LineNumbers>
{parsedContent.lineNumbers.map((lineNum, i) => (
<div key={i}>{lineNum}</div>
))}
</LineNumbers>
<ContentText>{parsedContent.actualContent}</ContentText>
</ContentBlock>
</DetailSection>
)}
</>
)}

Expand Down
6 changes: 5 additions & 1 deletion src/services/aiService.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -455,9 +455,13 @@ export class AIService extends EventEmitter {
log.debug_obj(`${workspaceId}/2a_redacted_messages.json`, redactedForProvider);

// Convert CmuxMessage to ModelMessage format using Vercel AI SDK utility
// Pass earlyTools so convertToModelMessages can use toModelOutput for tool results
// (earlyTools has stub config but same tool definitions with toModelOutput functions)
// Type assertion needed because CmuxMessage has custom tool parts for interrupted tools
// eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-argument
const modelMessages = convertToModelMessages(redactedForProvider as any);
const modelMessages = convertToModelMessages(redactedForProvider as any, {
tools: earlyTools,
});
log.debug_obj(`${workspaceId}/2_model_messages.json`, modelMessages);

// Apply ModelMessage transforms based on provider requirements
Expand Down
Loading