feat: adjust read tool so that it can handle dirs too (#13090)

This commit is contained in:
Aiden Cline
2026-02-11 13:23:00 -06:00
committed by GitHub
parent e3471526f4
commit 6b4d617df0
5 changed files with 115 additions and 22 deletions

View File

@@ -32,6 +32,9 @@ description: Use this when you are working on file operations like reading, writ
- Decode tool stderr with `Bun.readableStreamToText`.
- For large writes, use `Bun.write(Bun.file(path), text)`.
NOTE: Bun.file(...).exists() will return `false` if the value is a directory.
Use Filesystem.exists(...) instead if path can be file or directory
## Quick checklist
- Use Bun APIs first.

View File

@@ -2,9 +2,9 @@ Performs exact string replacements in files.
Usage:
- You must use your `Read` tool at least once in the conversation before editing. This tool will error if you attempt an edit without reading the file.
- When editing text from Read tool output, ensure you preserve the exact indentation (tabs/spaces) as it appears AFTER the line number prefix. The line number prefix format is: spaces + line number + tab. Everything after that tab is the actual file content to match. Never include any part of the line number prefix in the oldString or newString.
- When editing text from Read tool output, ensure you preserve the exact indentation (tabs/spaces) as it appears AFTER the line number prefix. The line number prefix format is: line number + colon + space (e.g., `1: `). Everything after that space is the actual file content to match. Never include any part of the line number prefix in the oldString or newString.
- ALWAYS prefer editing existing files in the codebase. NEVER write new files unless explicitly required.
- Only use emojis if the user explicitly requests it. Avoid adding emojis to files unless asked.
- The edit will FAIL if `oldString` is not found in the file with an error "oldString not found in content".
- The edit will FAIL if `oldString` is found multiple times in the file with an error "oldString found multiple times and requires more code context to uniquely identify the intended match". Either provide a larger string with more surrounding context to make it unique or use `replaceAll` to change every instance of `oldString`.
- The edit will FAIL if `oldString` is found multiple times in the file with an error "Found multiple matches for oldString. Provide more surrounding lines in oldString to identify the correct match." Either provide a larger string with more surrounding context to make it unique or use `replaceAll` to change every instance of `oldString`.
- Use `replaceAll` for replacing and renaming strings across the file. This parameter is useful if you want to rename a variable for instance.

View File

@@ -17,9 +17,9 @@ const MAX_BYTES = 50 * 1024
export const ReadTool = Tool.define("read", {
description: DESCRIPTION,
parameters: z.object({
filePath: z.string().describe("The path to the file to read"),
offset: z.coerce.number().describe("The line number to start reading from (0-based)").optional(),
limit: z.coerce.number().describe("The number of lines to read (defaults to 2000)").optional(),
filePath: z.string().describe("The absolute path to the file or directory to read"),
offset: z.coerce.number().describe("The 0-based line offset to start reading from").optional(),
limit: z.coerce.number().describe("The maximum number of lines to read (defaults to 2000)").optional(),
}),
async execute(params, ctx) {
let filepath = params.filePath
@@ -28,8 +28,12 @@ export const ReadTool = Tool.define("read", {
}
const title = path.relative(Instance.worktree, filepath)
const file = Bun.file(filepath)
const stat = await file.stat().catch(() => undefined)
await assertExternalDirectory(ctx, filepath, {
bypass: Boolean(ctx.extra?.["bypassCwdCheck"]),
kind: stat?.isDirectory() ? "directory" : "file",
})
await ctx.ask({
@@ -39,8 +43,7 @@ export const ReadTool = Tool.define("read", {
metadata: {},
})
const file = Bun.file(filepath)
if (!(await file.exists())) {
if (!stat) {
const dir = path.dirname(filepath)
const base = path.basename(filepath)
@@ -60,6 +63,47 @@ export const ReadTool = Tool.define("read", {
throw new Error(`File not found: ${filepath}`)
}
if (stat.isDirectory()) {
const dirents = await fs.promises.readdir(filepath, { withFileTypes: true })
const entries = await Promise.all(
dirents.map(async (dirent) => {
if (dirent.isDirectory()) return dirent.name + "/"
if (dirent.isSymbolicLink()) {
const target = await fs.promises.stat(path.join(filepath, dirent.name)).catch(() => undefined)
if (target?.isDirectory()) return dirent.name + "/"
}
return dirent.name
}),
)
entries.sort((a, b) => a.localeCompare(b))
const limit = params.limit ?? DEFAULT_READ_LIMIT
const offset = params.offset || 0
const sliced = entries.slice(offset, offset + limit)
const truncated = offset + sliced.length < entries.length
const output = [
`<path>${filepath}</path>`,
`<type>directory</type>`,
`<entries>`,
sliced.join("\n"),
truncated
? `\n(Showing ${sliced.length} of ${entries.length} entries. Use 'offset' parameter to read beyond entry ${offset + sliced.length})`
: `\n(${entries.length} entries)`,
`</entries>`,
].join("\n")
return {
title,
output,
metadata: {
preview: sliced.slice(0, 20).join("\n"),
truncated,
loaded: [] as string[],
},
}
}
const instructions = await InstructionPrompt.resolve(ctx.messages, filepath, ctx.messageID)
// Exclude SVG (XML-based) and vnd.fastbidsheet (.fbs extension, commonly FlatBuffers schema files)
@@ -75,7 +119,7 @@ export const ReadTool = Tool.define("read", {
metadata: {
preview: msg,
truncated: false,
...(instructions.length > 0 && { loaded: instructions.map((i) => i.filepath) }),
loaded: instructions.map((i) => i.filepath),
},
attachments: [
{
@@ -112,11 +156,11 @@ export const ReadTool = Tool.define("read", {
}
const content = raw.map((line, index) => {
return `${(index + offset + 1).toString().padStart(5, "0")}| ${line}`
return `${index + offset + 1}: ${line}`
})
const preview = raw.slice(0, 20).join("\n")
let output = "<file>\n"
let output = [`<path>${filepath}</path>`, `<type>file</type>`, "<content>"].join("\n")
output += content.join("\n")
const totalLines = lines.length
@@ -131,7 +175,7 @@ export const ReadTool = Tool.define("read", {
} else {
output += `\n\n(End of file - total ${totalLines} lines)`
}
output += "\n</file>"
output += "\n</content>"
// just warms the lsp client
LSP.touchFile(filepath, false)
@@ -147,7 +191,7 @@ export const ReadTool = Tool.define("read", {
metadata: {
preview,
truncated,
...(instructions.length > 0 && { loaded: instructions.map((i) => i.filepath) }),
loaded: instructions.map((i) => i.filepath),
},
}
},

View File

@@ -1,12 +1,13 @@
Reads a file from the local filesystem. You can access any file directly by using this tool.
Assume this tool is able to read all files on the machine. If the User provides a path to a file assume that path is valid. It is okay to read a file that does not exist; an error will be returned.
Read a file or directory from the local filesystem. If the path does not exist, an error is returned.
Usage:
- The filePath parameter must be an absolute path, not a relative path
- By default, it reads up to 2000 lines starting from the beginning of the file
- You can optionally specify a line offset and limit (especially handy for long files), but it's recommended to read the whole file by not providing these parameters
- Any lines longer than 2000 characters will be truncated
- Results are returned using cat -n format, with line numbers starting at 1
- You have the capability to call multiple tools in a single response. It is always better to speculatively read multiple files as a batch that are potentially useful.
- If you read a file that exists but has empty contents you will receive a system reminder warning in place of file contents.
- You can read image files using this tool.
- The filePath parameter should be an absolute path.
- By default, this tool returns up to 2000 lines from the start of the file.
- To read later sections, call this tool again with a larger offset.
- Use the grep tool to find specific content in large files or files with long lines.
- If you are unsure of the correct file path, use the glob tool to look up filenames by glob pattern.
- Contents are returned with each line prefixed by its line number as `<line>: <content>`. For example, if a file has contents "foo\n", you will receive "1: foo\n". For directories, entries are returned one per line (without line numbers) with a trailing `/` for subdirectories.
- Any line longer than 2000 characters is truncated.
- Call this tool in parallel when you know there are multiple files you want to read.
- Avoid tiny repeated slices (30 line chunks). If you need more context, read a larger window.
- This tool can read image files and PDFs and return them as file attachments.

View File

@@ -78,6 +78,32 @@ describe("tool.read external_directory permission", () => {
})
})
test("asks for directory-scoped external_directory permission when reading external directory", async () => {
await using outerTmp = await tmpdir({
init: async (dir) => {
await Bun.write(path.join(dir, "external", "a.txt"), "a")
},
})
await using tmp = await tmpdir({ git: true })
await Instance.provide({
directory: tmp.path,
fn: async () => {
const read = await ReadTool.init()
const requests: Array<Omit<PermissionNext.Request, "id" | "sessionID" | "tool">> = []
const testCtx = {
...ctx,
ask: async (req: Omit<PermissionNext.Request, "id" | "sessionID" | "tool">) => {
requests.push(req)
},
}
await read.execute({ filePath: path.join(outerTmp.path, "external") }, testCtx)
const extDirReq = requests.find((r) => r.permission === "external_directory")
expect(extDirReq).toBeDefined()
expect(extDirReq!.patterns).toContain(path.join(outerTmp.path, "external", "*"))
},
})
})
test("asks for external_directory permission when reading relative path outside project", async () => {
await using tmp = await tmpdir({ git: true })
await Instance.provide({
@@ -249,6 +275,25 @@ describe("tool.read truncation", () => {
})
})
test("does not mark final directory page as truncated", async () => {
await using tmp = await tmpdir({
init: async (dir) => {
await Promise.all(
Array.from({ length: 10 }, (_, i) => Bun.write(path.join(dir, "dir", `file-${i}.txt`), `line${i}`)),
)
},
})
await Instance.provide({
directory: tmp.path,
fn: async () => {
const read = await ReadTool.init()
const result = await read.execute({ filePath: path.join(tmp.path, "dir"), offset: 5, limit: 5 }, ctx)
expect(result.metadata.truncated).toBe(false)
expect(result.output).not.toContain("Showing 5 of 10 entries")
},
})
})
test("truncates long lines", async () => {
await using tmp = await tmpdir({
init: async (dir) => {