@constructive-db/machine-protocol 0.2.2 → 0.4.0

package/esm/index.js

@@ -236,6 +236,53 @@ export class LineSplitter {
 function stripCarriageReturn(line) {
     return line.endsWith('\r') ? line.slice(0, -1) : line;
 }
+/**
+ * Reading paths under a root, read-only. Two vocabularies, like everything
+ * else: the runner leg (`path_list` / `path_read`) names a root relative to the
+ * runner's policy cwd and knows nothing about sessions; the client leg
+ * (`fs_list` / `fs_read`) names a session, and the relay maps it to that
+ * session's working tree after the tenant authorizes it. A request carries a
+ * `requestId` and is answered by exactly one frame with the same id: a
+ * listing, a content frame, or an `error` with a {@link PathErrorCode}.
+ */
+/** The most bytes one read returns. A larger file is refused, never truncated. */
+export const MAX_READ_BYTES = 1024 * 1024;
+/** The most entries one listing returns. A larger directory is refused, never truncated. */
+export const MAX_LIST_ENTRIES = 1000;
+/** Path requests a runner answers at once; one more is refused, not queued. */
+export const MAX_PATH_REQUESTS_IN_FLIGHT = 8;
+export const PATH_ERROR_CODES = [
+    /** Absolute, traversing (`..`), or carrying a NUL byte. */
+    'PATH_INVALID',
+    /** Resolves — through a symlink — outside the root. */
+    'PATH_OUTSIDE_ROOT',
+    'PATH_NOT_FOUND',
+    'NOT_A_DIRECTORY',
+    'NOT_A_FILE',
+    'FILE_TOO_LARGE',
+    'DIRECTORY_TOO_LARGE',
+    /** More than {@link MAX_PATH_REQUESTS_IN_FLIGHT} requests are being answered. */
+    'TOO_MANY_PATH_REQUESTS',
+    /** Any other filesystem failure; the message carries the cause. */
+    'PATH_UNREADABLE'
+];
+/**
+ * Why `requested` cannot name a path under a root, or `null` when it can.
+ * `''` (and `.`) name the root itself. This is syntax only: a path that passes
+ * may still resolve outside the root through a symlink, which only the side
+ * holding the filesystem can tell.
+ */
+export function relativePathProblem(requested) {
+    if (requested.includes('\0'))
+        return 'contains a NUL byte';
+    if (requested.startsWith('/') || requested.startsWith('\\') || /^[A-Za-z]:/.test(requested)) {
+        return 'is absolute';
+    }
+    if (requested.split(/[\\/]/).includes('..'))
+        return "traverses out of its root with '..'";
+    return null;
+}
+const PATH_ENTRY_KINDS = new Set(['file', 'directory', 'symlink', 'other']);
 const FRAME_TYPES = new Set([
     'open',
     'input',
@@ -252,7 +299,15 @@ const FRAME_TYPES = new Set([
     'attached',
     'detached',
     'reattached',
-    'enrolled'
+    'enrolled',
+    'path_list',
+    'path_read',
+    'path_listing',
+    'path_content',
+    'fs_list',
+    'fs_read',
+    'fs_listing',
+    'fs_content'
 ]);
 /** Frame types that must carry a session id. */
 const SESSION_FRAME_TYPES = new Set([
@@ -268,8 +323,31 @@ const SESSION_FRAME_TYPES = new Set([
     'exit',
     'scrollback',
     'detached',
-    'reattached'
+    'reattached',
+    'fs_list',
+    'fs_read',
+    'fs_listing',
+    'fs_content'
+]);
+/** Frame types that answer, or ask, one request. */
+const REQUEST_FRAME_TYPES = new Set([
+    'path_list',
+    'path_read',
+    'path_listing',
+    'path_content',
+    'fs_list',
+    'fs_read',
+    'fs_listing',
+    'fs_content'
 ]);
+function validEntries(value) {
+    return (Array.isArray(value) &&
+        value.every(entry => isPlainObject(entry) &&
+            typeof entry.name === 'string' &&
+            typeof entry.kind === 'string' &&
+            PATH_ENTRY_KINDS.has(entry.kind) &&
+            (entry.size === undefined || (Number.isInteger(entry.size) && entry.size >= 0))));
+}
 const SIGNALS = new Set(SIGNAL_NAMES);
 /**
  * Terminal geometry a pty will accept. A resize is a client-supplied pair of
@@ -410,6 +488,10 @@ export function decodeFrame(raw) {
     if (SESSION_FRAME_TYPES.has(type) && typeof frame.sessionId !== 'string') {
         throw new Error(`machine-protocol: '${type}' frame is missing sessionId`);
     }
+    if (REQUEST_FRAME_TYPES.has(type) &&
+        (typeof frame.requestId !== 'string' || frame.requestId.length === 0)) {
+        throw new Error(`machine-protocol: '${type}' frame is missing requestId`);
+    }
     switch (type) {
         case 'open':
             if (typeof frame.command !== 'string' || frame.command.length === 0) {
@@ -475,6 +557,48 @@ export function decodeFrame(raw) {
             if (typeof frame.message !== 'string') {
                 throw new Error("machine-protocol: 'error' frame is missing message");
             }
+            if (frame.requestId !== undefined && typeof frame.requestId !== 'string') {
+                throw new Error("machine-protocol: 'error' frame requestId must be a string");
+            }
+            if (frame.code !== undefined && typeof frame.code !== 'string') {
+                throw new Error("machine-protocol: 'error' frame code must be a string");
+            }
+            break;
+        case 'path_list':
+        case 'path_read':
+            if (frame.root !== undefined && typeof frame.root !== 'string') {
+                throw new Error(`machine-protocol: '${type}' frame root must be a string`);
+            }
+            if (typeof frame.path !== 'string') {
+                throw new Error(`machine-protocol: '${type}' frame is missing path`);
+            }
+            break;
+        case 'fs_list':
+        case 'fs_read':
+            if (typeof frame.path !== 'string') {
+                throw new Error(`machine-protocol: '${type}' frame is missing path`);
+            }
+            break;
+        case 'path_listing':
+        case 'fs_listing':
+            if (!validEntries(frame.entries)) {
+                throw new Error(`machine-protocol: '${type}' frame has malformed entries`);
+            }
+            if (type === 'fs_listing' && typeof frame.path !== 'string') {
+                throw new Error("machine-protocol: 'fs_listing' frame is missing path");
+            }
+            break;
+        case 'path_content':
+        case 'fs_content':
+            if (typeof frame.data !== 'string' || !Number.isInteger(frame.size)) {
+                throw new Error(`machine-protocol: '${type}' frame is missing data/size`);
+            }
+            if (frame.size !== Buffer.from(frame.data, 'base64').length) {
+                throw new Error(`machine-protocol: '${type}' frame size does not match its data`);
+            }
+            if (type === 'fs_content' && typeof frame.path !== 'string') {
+                throw new Error("machine-protocol: 'fs_content' frame is missing path");
+            }
             break;
         case 'attached':
             if (typeof frame.machineId !== 'string') {

package/index.d.ts

@@ -265,10 +265,103 @@ export interface ExitFrame {
 export interface ErrorFrame {
     type: 'error';
     sessionId?: string;
+    /** The `path_*` / `fs_*` request this error answers, when it answers one. */
+    requestId?: string;
     message: string;
     /** A stable, machine-readable reason, where the sender has one. */
     code?: string;
 }
+/**
+ * Reading paths under a root, read-only. Two vocabularies, like everything
+ * else: the runner leg (`path_list` / `path_read`) names a root relative to the
+ * runner's policy cwd and knows nothing about sessions; the client leg
+ * (`fs_list` / `fs_read`) names a session, and the relay maps it to that
+ * session's working tree after the tenant authorizes it. A request carries a
+ * `requestId` and is answered by exactly one frame with the same id: a
+ * listing, a content frame, or an `error` with a {@link PathErrorCode}.
+ */
+/** The most bytes one read returns. A larger file is refused, never truncated. */
+export declare const MAX_READ_BYTES: number;
+/** The most entries one listing returns. A larger directory is refused, never truncated. */
+export declare const MAX_LIST_ENTRIES = 1000;
+/** Path requests a runner answers at once; one more is refused, not queued. */
+export declare const MAX_PATH_REQUESTS_IN_FLIGHT = 8;
+export declare const PATH_ERROR_CODES: readonly ["PATH_INVALID", "PATH_OUTSIDE_ROOT", "PATH_NOT_FOUND", "NOT_A_DIRECTORY", "NOT_A_FILE", "FILE_TOO_LARGE", "DIRECTORY_TOO_LARGE", "TOO_MANY_PATH_REQUESTS", "PATH_UNREADABLE"];
+export type PathErrorCode = (typeof PATH_ERROR_CODES)[number];
+/**
+ * Why `requested` cannot name a path under a root, or `null` when it can.
+ * `''` (and `.`) name the root itself. This is syntax only: a path that passes
+ * may still resolve outside the root through a symlink, which only the side
+ * holding the filesystem can tell.
+ */
+export declare function relativePathProblem(requested: string): string | null;
+export type PathEntryKind = 'file' | 'directory' | 'symlink' | 'other';
+export interface PathEntry {
+    name: string;
+    /** What the entry itself is; a symlink is reported as one, not followed. */
+    kind: PathEntryKind;
+    /** Bytes, for a file. */
+    size?: number;
+}
+/** Relay → runner: list a directory under `root` (relative to the policy cwd). */
+export interface PathListFrame {
+    type: 'path_list';
+    requestId: string;
+    /** The root `path` is confined to; omitted means the policy cwd itself. */
+    root?: string;
+    path: string;
+}
+/** Relay → runner: read a file under `root` (relative to the policy cwd). */
+export interface PathReadFrame {
+    type: 'path_read';
+    requestId: string;
+    root?: string;
+    path: string;
+}
+/** Runner → relay: the entries of a `path_list`, sorted by name. */
+export interface PathListingFrame {
+    type: 'path_listing';
+    requestId: string;
+    entries: PathEntry[];
+}
+/** Runner → relay: a `path_read`'s bytes, base64. */
+export interface PathContentFrame {
+    type: 'path_content';
+    requestId: string;
+    data: string;
+    size: number;
+}
+/** Client → relay: list a directory in a session's working tree. */
+export interface ListFilesFrame {
+    type: 'fs_list';
+    sessionId: string;
+    requestId: string;
+    path: string;
+}
+/** Client → relay: read a file in a session's working tree. */
+export interface ReadFileFrame {
+    type: 'fs_read';
+    sessionId: string;
+    requestId: string;
+    path: string;
+}
+/** Relay → client: the answer to an `fs_list`. */
+export interface FileListingFrame {
+    type: 'fs_listing';
+    sessionId: string;
+    requestId: string;
+    path: string;
+    entries: PathEntry[];
+}
+/** Relay → client: the answer to an `fs_read`, base64. */
+export interface FileContentFrame {
+    type: 'fs_content';
+    sessionId: string;
+    requestId: string;
+    path: string;
+    data: string;
+    size: number;
+}
 /**
  * Runner → relay: the scrollback a detaching session leaves behind, capped by
  * the runner. The relay records it as the session's one interactive output row
@@ -325,9 +418,9 @@ export interface EnrolledFrame {
     /** Epoch seconds at which {@link credential} stops verifying. */
     expiresAt: number;
 }
-export type ClientToRunnerFrame = OpenFrame | InputFrame | ResizeFrame | SignalFrame | DetachFrame | ReattachFrame | CloseFrame;
-export type RunnerToClientFrame = OutputFrame | AgentEventFrame | ExitFrame | ErrorFrame | ScrollbackFrame;
-export type Frame = ClientToRunnerFrame | RunnerToClientFrame | AttachedFrame | DetachedFrame | ReattachedFrame | EnrolledFrame;
+export type ClientToRunnerFrame = OpenFrame | InputFrame | ResizeFrame | SignalFrame | DetachFrame | ReattachFrame | CloseFrame | PathListFrame | PathReadFrame;
+export type RunnerToClientFrame = OutputFrame | AgentEventFrame | ExitFrame | ErrorFrame | ScrollbackFrame | PathListingFrame | PathContentFrame;
+export type Frame = ClientToRunnerFrame | RunnerToClientFrame | ListFilesFrame | ReadFileFrame | FileListingFrame | FileContentFrame | AttachedFrame | DetachedFrame | ReattachedFrame | EnrolledFrame;
 /**
  * Whether `value` looks like an {@link AgentEvent} — an object naming one of
  * the event kinds. Shape is not checked; that is {@link assertAgentEvent}. A

package/index.js

@@ -36,12 +36,13 @@
 // the tenant that issued it, and RLS decides everything after that. Addressing a
 // database you have no credential for is simply a refusal.
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.LineSplitter = exports.MAX_LINE_LENGTH = exports.AGENT_CLI_COMMAND = exports.AGENT_EVENT_KINDS = exports.SIGNAL_NAMES = exports.AGENT_MODES = exports.attachSubprotocols = exports.ATTACH_SUBPROTOCOL = exports.clientAttachPath = exports.RUNNER_CONNECT_PATH = exports.HEADER_MACHINE_DATABASE = exports.HEADER_MACHINE_ID = void 0;
+exports.PATH_ERROR_CODES = exports.MAX_PATH_REQUESTS_IN_FLIGHT = exports.MAX_LIST_ENTRIES = exports.MAX_READ_BYTES = exports.LineSplitter = exports.MAX_LINE_LENGTH = exports.AGENT_CLI_COMMAND = exports.AGENT_EVENT_KINDS = exports.SIGNAL_NAMES = exports.AGENT_MODES = exports.attachSubprotocols = exports.ATTACH_SUBPROTOCOL = exports.clientAttachPath = exports.RUNNER_CONNECT_PATH = exports.HEADER_MACHINE_DATABASE = exports.HEADER_MACHINE_ID = void 0;
 exports.credentialFromSubprotocols = credentialFromSubprotocols;
 exports.parseClientAttachPath = parseClientAttachPath;
 exports.agentBinding = agentBinding;
 exports.encodeApprovalDecisionLine = encodeApprovalDecisionLine;
 exports.parseApprovalDecisionLine = parseApprovalDecisionLine;
+exports.relativePathProblem = relativePathProblem;
 exports.isAgentEventLike = isAgentEventLike;
 exports.assertAgentEvent = assertAgentEvent;
 exports.encodeFrame = encodeFrame;
@@ -251,6 +252,53 @@ exports.LineSplitter = LineSplitter;
 function stripCarriageReturn(line) {
     return line.endsWith('\r') ? line.slice(0, -1) : line;
 }
+/**
+ * Reading paths under a root, read-only. Two vocabularies, like everything
+ * else: the runner leg (`path_list` / `path_read`) names a root relative to the
+ * runner's policy cwd and knows nothing about sessions; the client leg
+ * (`fs_list` / `fs_read`) names a session, and the relay maps it to that
+ * session's working tree after the tenant authorizes it. A request carries a
+ * `requestId` and is answered by exactly one frame with the same id: a
+ * listing, a content frame, or an `error` with a {@link PathErrorCode}.
+ */
+/** The most bytes one read returns. A larger file is refused, never truncated. */
+exports.MAX_READ_BYTES = 1024 * 1024;
+/** The most entries one listing returns. A larger directory is refused, never truncated. */
+exports.MAX_LIST_ENTRIES = 1000;
+/** Path requests a runner answers at once; one more is refused, not queued. */
+exports.MAX_PATH_REQUESTS_IN_FLIGHT = 8;
+exports.PATH_ERROR_CODES = [
+    /** Absolute, traversing (`..`), or carrying a NUL byte. */
+    'PATH_INVALID',
+    /** Resolves — through a symlink — outside the root. */
+    'PATH_OUTSIDE_ROOT',
+    'PATH_NOT_FOUND',
+    'NOT_A_DIRECTORY',
+    'NOT_A_FILE',
+    'FILE_TOO_LARGE',
+    'DIRECTORY_TOO_LARGE',
+    /** More than {@link MAX_PATH_REQUESTS_IN_FLIGHT} requests are being answered. */
+    'TOO_MANY_PATH_REQUESTS',
+    /** Any other filesystem failure; the message carries the cause. */
+    'PATH_UNREADABLE'
+];
+/**
+ * Why `requested` cannot name a path under a root, or `null` when it can.
+ * `''` (and `.`) name the root itself. This is syntax only: a path that passes
+ * may still resolve outside the root through a symlink, which only the side
+ * holding the filesystem can tell.
+ */
+function relativePathProblem(requested) {
+    if (requested.includes('\0'))
+        return 'contains a NUL byte';
+    if (requested.startsWith('/') || requested.startsWith('\\') || /^[A-Za-z]:/.test(requested)) {
+        return 'is absolute';
+    }
+    if (requested.split(/[\\/]/).includes('..'))
+        return "traverses out of its root with '..'";
+    return null;
+}
+const PATH_ENTRY_KINDS = new Set(['file', 'directory', 'symlink', 'other']);
 const FRAME_TYPES = new Set([
     'open',
     'input',
@@ -267,7 +315,15 @@ const FRAME_TYPES = new Set([
     'attached',
     'detached',
     'reattached',
-    'enrolled'
+    'enrolled',
+    'path_list',
+    'path_read',
+    'path_listing',
+    'path_content',
+    'fs_list',
+    'fs_read',
+    'fs_listing',
+    'fs_content'
 ]);
 /** Frame types that must carry a session id. */
 const SESSION_FRAME_TYPES = new Set([
@@ -283,8 +339,31 @@ const SESSION_FRAME_TYPES = new Set([
     'exit',
     'scrollback',
     'detached',
-    'reattached'
+    'reattached',
+    'fs_list',
+    'fs_read',
+    'fs_listing',
+    'fs_content'
+]);
+/** Frame types that answer, or ask, one request. */
+const REQUEST_FRAME_TYPES = new Set([
+    'path_list',
+    'path_read',
+    'path_listing',
+    'path_content',
+    'fs_list',
+    'fs_read',
+    'fs_listing',
+    'fs_content'
 ]);
+function validEntries(value) {
+    return (Array.isArray(value) &&
+        value.every(entry => isPlainObject(entry) &&
+            typeof entry.name === 'string' &&
+            typeof entry.kind === 'string' &&
+            PATH_ENTRY_KINDS.has(entry.kind) &&
+            (entry.size === undefined || (Number.isInteger(entry.size) && entry.size >= 0))));
+}
 const SIGNALS = new Set(exports.SIGNAL_NAMES);
 /**
  * Terminal geometry a pty will accept. A resize is a client-supplied pair of
@@ -425,6 +504,10 @@ function decodeFrame(raw) {
     if (SESSION_FRAME_TYPES.has(type) && typeof frame.sessionId !== 'string') {
         throw new Error(`machine-protocol: '${type}' frame is missing sessionId`);
     }
+    if (REQUEST_FRAME_TYPES.has(type) &&
+        (typeof frame.requestId !== 'string' || frame.requestId.length === 0)) {
+        throw new Error(`machine-protocol: '${type}' frame is missing requestId`);
+    }
     switch (type) {
         case 'open':
             if (typeof frame.command !== 'string' || frame.command.length === 0) {
@@ -490,6 +573,48 @@ function decodeFrame(raw) {
             if (typeof frame.message !== 'string') {
                 throw new Error("machine-protocol: 'error' frame is missing message");
             }
+            if (frame.requestId !== undefined && typeof frame.requestId !== 'string') {
+                throw new Error("machine-protocol: 'error' frame requestId must be a string");
+            }
+            if (frame.code !== undefined && typeof frame.code !== 'string') {
+                throw new Error("machine-protocol: 'error' frame code must be a string");
+            }
+            break;
+        case 'path_list':
+        case 'path_read':
+            if (frame.root !== undefined && typeof frame.root !== 'string') {
+                throw new Error(`machine-protocol: '${type}' frame root must be a string`);
+            }
+            if (typeof frame.path !== 'string') {
+                throw new Error(`machine-protocol: '${type}' frame is missing path`);
+            }
+            break;
+        case 'fs_list':
+        case 'fs_read':
+            if (typeof frame.path !== 'string') {
+                throw new Error(`machine-protocol: '${type}' frame is missing path`);
+            }
+            break;
+        case 'path_listing':
+        case 'fs_listing':
+            if (!validEntries(frame.entries)) {
+                throw new Error(`machine-protocol: '${type}' frame has malformed entries`);
+            }
+            if (type === 'fs_listing' && typeof frame.path !== 'string') {
+                throw new Error("machine-protocol: 'fs_listing' frame is missing path");
+            }
+            break;
+        case 'path_content':
+        case 'fs_content':
+            if (typeof frame.data !== 'string' || !Number.isInteger(frame.size)) {
+                throw new Error(`machine-protocol: '${type}' frame is missing data/size`);
+            }
+            if (frame.size !== Buffer.from(frame.data, 'base64').length) {
+                throw new Error(`machine-protocol: '${type}' frame size does not match its data`);
+            }
+            if (type === 'fs_content' && typeof frame.path !== 'string') {
+                throw new Error("machine-protocol: 'fs_content' frame is missing path");
+            }
             break;
         case 'attached':
             if (typeof frame.machineId !== 'string') {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@constructive-db/machine-protocol",
-  "version": "0.2.2",
+  "version": "0.4.0",
   "description": "Wire protocol shared by the machine relay and the machine runner: the JSON frames a session is made of, and the codec that refuses malformed ones.",
   "author": "Constructive <developers@constructive.io>",
   "license": "MIT",
@@ -28,5 +28,5 @@
     "lint": "eslint . --fix",
     "test": "jest"
   },
-  "gitHead": "81eed19332cc231de0fde02bc6e87002357f9536"
+  "gitHead": "487c4ec9ee997d7da80584a3d41d7120d4354c54"
 }