@evident-ai/runner-cdk 3.5.2-dev.8d3aa06 → 3.5.2-dev.bbd799b

package/dist/microvm-image-context/Dockerfile

@@ -1,190 +0,0 @@
-# Generic MicroVM runner image
-#
-# Built for linux/arm64 (Lambda MicroVMs run on Graviton only) and with ONE
-# process: the hook server. `overlay/setup-root` and `overlay/setup-workspace`
-# are the extension point for deployment-specific packages, dependency
-# installation and build warm-up.
-#
-# THE BOOT SPLIT — the rule that governs everything below. AWS snapshots this
-# image ONCE at build and every MicroVM resumes from that same snapshot, so
-# anything unique or secret baked in here is shared by every VM launched from
-# this image version. Therefore this image must NOT start the tunnel
-# (`evident run`), clone the repo, or carry any credential, machine ID or
-# generated secret. Identity arrives per-VM in the `/run` hook payload.
-# `aws/runner-cdk/src/microvm/image/dockerfile.test.ts` fails the build if this
-# regresses.
-#
-# No `--platform` on FROM: AWS builds this natively on Graviton. Build it
-# locally with `docker build --platform linux/arm64`.
-#
-# Node major tracks `.nvmrc` / the root `engines.node` (22) so the image runs
-# the same runtime as CI — `evident-runner/src/node-version.test.ts` fails if
-# they drift. The Debian codename is pinned explicitly (`bookworm`, what
-# `node:22-slim` resolves to today) so a future default bump to a newer Debian
-# can't silently change the apt package set this image asserts on below.
-FROM node:22-bookworm-slim
-
-ENV PNPM_HOME="/pnpm"
-ENV PATH="$PNPM_HOME:$PATH"
-RUN corepack enable && corepack prepare pnpm@9.15.0 --activate
-
-# System tooling + GitHub CLI + AWS CLI, in a single layer.
-#   - awscli (v2): required — durable-state persistence uses `aws s3`.
-#   - procps: required at RUNTIME for process inspection and debugging. Listed
-#     explicitly so apt keeps it.
-#   - gnupg, unzip: BUILD-ONLY (gh apt key / AWS CLI bundle); purged in this layer.
-#
-# Purge build-only packages BY NAME — do NOT use `apt-get autoremove`, which
-# once cascaded into shared libs and swept `procps` out of the runner image. The
-# binary assertion further down fails the build if any required binary is gone.
-RUN apt-get update \
-  && apt-get install -y --no-install-recommends \
-    ca-certificates curl git jq less procps psmisc ripgrep \
-    gnupg unzip \
-  # --- GitHub CLI (`gh`) from the official apt repo ---
-  && mkdir -p -m 755 /etc/apt/keyrings \
-  && curl -fsSL https://cli.github.com/packages/githubcli-archive-keyring.gpg \
-       -o /etc/apt/keyrings/githubcli-archive-keyring.gpg \
-  && chmod go+r /etc/apt/keyrings/githubcli-archive-keyring.gpg \
-  && echo "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/githubcli-archive-keyring.gpg] https://cli.github.com/packages stable main" \
-       > /etc/apt/sources.list.d/github-cli.list \
-  && apt-get update \
-  && apt-get install -y --no-install-recommends gh \
-  # --- AWS CLI v2 (aarch64 bundle; this image is arm64 only) ---
-  && curl -fsSL "https://awscli.amazonaws.com/awscli-exe-linux-aarch64.zip" -o /tmp/awscliv2.zip \
-  && unzip -q /tmp/awscliv2.zip -d /tmp \
-  && /tmp/aws/install \
-  && aws --version \
-  # --- Drop the build-only packages (BY NAME, no autoremove) + caches ---
-  && apt-get purge -y gnupg unzip \
-  && rm -rf /tmp/aws /tmp/awscliv2.zip /var/lib/apt/lists/*
-
-# Litestream — replicates the OpenCode SQLite session DB to S3 so history
-# survives VM replacement. Pinned + checksum-verified: a wrong LITESTREAM_SHA256
-# fails the build at `sha256sum -c`. The checksum is for the arm64 tarball and
-# is NOT interchangeable with the ECS runner image's. Note the release tag keeps
-# the `v` but the filename does not.
-# `/run` restores opencode.db from S3 (#812 WI-2) — one bounded `litestream
-# restore` attempt — then starts a backgrounded `litestream replicate` once
-# opencode has opened the DB (WI-3); `/suspend`/`/resume`/`/terminate` flush,
-# restart and stop it in turn (WI-4). Do not remove as dead code.
-ARG LITESTREAM_VERSION=0.5.16
-ARG LITESTREAM_SHA256=678022e4103145302598e35d37f8718392d42e153feeb1e2d4a64dd0cd3aaf10
-RUN curl -fsSL -o /tmp/litestream.tar.gz \
-      "https://github.com/benbjohnson/litestream/releases/download/v${LITESTREAM_VERSION}/litestream-${LITESTREAM_VERSION}-linux-arm64.tar.gz" \
-  && echo "${LITESTREAM_SHA256}  /tmp/litestream.tar.gz" | sha256sum -c - \
-  && tar -C /usr/local/bin -xzf /tmp/litestream.tar.gz litestream \
-  && chmod +x /usr/local/bin/litestream \
-  && rm -f /tmp/litestream.tar.gz \
-  && litestream version
-
-# The AI agent runtime, the Evident CLI and the Claude Code CLI, installed
-# globally and pinned via build args so image versions are reproducible.
-# OPENCODE_VERSION is a DELIBERATE PIN, not a floating tag — see #537: a
-# floating `opencode-ai@latest` install was implicated in unbounded runner
-# memory growth / OOM kills. Keep it in lockstep with the ECS runner image's.
-# EVIDENT_CLI_VERSION is DELIBERATELY the floating `dev` tag, matching the ECS
-# runner image. This is a
-# development environment: it should track the head of the CLI so changes are
-# testable without a version bump in every PR that touches a hook.
-# RUNNER_SYNCHRONISER_VERSION floats on `dev` for the identical reason.
-ARG OPENCODE_VERSION=1.18.3
-ARG EVIDENT_CLI_VERSION=dev
-ARG CLAUDE_CODE_VERSION=latest
-ARG RUNNER_SYNCHRONISER_VERSION=dev
-RUN npm install -g \
-    "opencode-ai@${OPENCODE_VERSION}" \
-    "@evident-ai/cli@${EVIDENT_CLI_VERSION}" \
-    "@anthropic-ai/claude-code@${CLAUDE_CODE_VERSION}" \
-    "@evident-ai/runner-synchroniser@${RUNNER_SYNCHRONISER_VERSION}" \
-  && opencode --version \
-  && evident --version \
-  && npm cache clean --force
-
-# Assert every REQUIRED runtime binary survived every layer above (fail loud).
-# Placed after the last install so it covers the apt purge, the litestream
-# extraction and the global npm installs in one pass.
-RUN for bin in ps pgrep pkill curl aws gh git jq rg litestream runner-synchroniser opencode evident timeout; do \
-       command -v "$bin" >/dev/null 2>&1 || { echo "FATAL: required binary '$bin' missing from image" >&2; exit 1; }; \
-     done \
-  && curl -fsS -o /dev/null https://cli.github.com \
-     || { echo "FATAL: curl TLS smoke test failed (broken TLS libs)" >&2; exit 1; }
-
-# Non-root user that owns the workspace. uid/gid pinned to 10001 for stable file
-# ownership (opencode.db, ~/.claude/.credentials.json) across rebuilds.
-RUN groupadd --system --gid 10001 runner \
-  && useradd --system --uid 10001 --gid 10001 \
-       --create-home --home-dir /home/runner runner
-
-# The hook server: the ONLY process this image starts, and it starts nothing
-# itself until AWS calls a hook. It is an esbuild bundle of
-# src/microvm/image/hook-server.ts, placed next to this Dockerfile by
-# `stageMicrovmImageContext()` — which is why this file is not buildable
-# straight from the source tree.
-COPY hook-server.js /usr/local/lib/evident/hook-server.js
-
-# One executable per lifecycle phase. There is deliberately no `ready` and no
-# `validate` script: the snapshot is shared by every VM started from this image
-# version, so nothing per-VM may run before /run — which leaves those two phases
-# with nothing to do, and the runtime answers them 200 without running anything.
-COPY hooks /etc/evident/hooks
-
-# Workspace: the repository the agent works on, baked in. A resumed VM must be
-# useful in milliseconds, while a clone costs minutes — so the checkout happens
-# here; dependency installation and build warm-up belong to the deployment overlay.
-#
-# This does not weaken the boot split: source, git history and node_modules are
-# neither secret nor per-VM-unique, so sharing them across every VM from this
-# snapshot is exactly right. No credential is involved either — the repo travels
-# in the build context (cloned by `stageMicrovmImageContext()` from a local
-# checkout), not through a `git clone` needing a token. `origin` is the plain
-# HTTPS GitHub URL; the credential that makes it fetchable and pushable arrives
-# per-VM, so the agent is what brings this checkout up to date.
-ENV WORKSPACE=/workspace
-ENV HOME=/home/runner
-#
-# The machine-id files are created EMPTY and runner-owned so that /run can
-# actually rewrite them: the hooks run as uid 10001, and neither `/etc` nor
-# `/var/lib/dbus` is writable by it, so without this the regeneration is a
-# permanent no-op and every VM from this snapshot shares one machine id. Empty
-# is the correct unset state — it is the absence of an identity, so nothing
-# per-VM-unique enters the shared snapshot.
-RUN mkdir -p /var/lib/dbus /home/runner/.local/state/evident \
-  && install -o runner -g runner -m 0644 /dev/null /etc/machine-id \
-  && install -o runner -g runner -m 0644 /dev/null /var/lib/dbus/machine-id \
-  && chown -R runner:runner /home/runner
-
-# Deployment additions are always present because staging writes no-op defaults
-# when no overlay is supplied.
-COPY overlay /etc/evident/overlay
-# HOME is already /home/runner, so root-warmed npx/npm state must be runner-owned
-# or the runner's first npm/pnpm call would fail with EACCES.
-RUN /etc/evident/overlay/setup-root \
-  && chown -R runner:runner /home/runner
-COPY --chown=10001:10001 repo ${WORKSPACE}
-WORKDIR ${WORKSPACE}
-USER runner
-
-# The staged context deliberately ships no `.git/index` (it is not byte-stable,
-# so shipping it would re-roll the image version on every deploy) — without this
-# reset the agent's workspace reports ~1400 files as staged deletions. Runs as
-# `runner` (not root) so the rebuilt index is runner-owned and writable later.
-RUN git reset --quiet
-
-# Deployment-specific dependency installation and build work belongs here. This runs
-# as `runner` in `${WORKSPACE}`, so everything it writes is agent-writable.
-#
-# Keep the overlay's install and build in one layer: pnpm hard-links `node_modules`
-# into its content-addressed store, and hard links only survive within a single layer.
-RUN /etc/evident/overlay/setup-workspace
-
-# Port AWS calls the MicroVM hooks on.
-ENV HOOKS_PORT=8080
-# OpenCode loopback port the CLI tunnels to (matches `evident run` default).
-ENV OPENCODE_PORT=4096
-
-EXPOSE 8080
-
-# No ENTRYPOINT shell: the hook server is started directly, and it must be the
-# only thing running until AWS calls /run.
-CMD ["node", "/usr/local/lib/evident/hook-server.js"]

package/dist/microvm-image-context/hook-server.js

@@ -1,286 +0,0 @@
-"use strict";
-var __create = Object.create;
-var __defProp = Object.defineProperty;
-var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
-var __getOwnPropNames = Object.getOwnPropertyNames;
-var __getProtoOf = Object.getPrototypeOf;
-var __hasOwnProp = Object.prototype.hasOwnProperty;
-var __commonJS = (cb, mod) => function __require() {
-  return mod || (0, cb[__getOwnPropNames(cb)[0]])((mod = { exports: {} }).exports, mod), mod.exports;
-};
-var __copyProps = (to, from, except, desc) => {
-  if (from && typeof from === "object" || typeof from === "function") {
-    for (let key of __getOwnPropNames(from))
-      if (!__hasOwnProp.call(to, key) && key !== except)
-        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
-  }
-  return to;
-};
-var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__getProtoOf(mod)) : {}, __copyProps(
-  // If the importer is in node compatibility mode or this is not an ESM
-  // file that has been converted to a CommonJS file using a Babel-
-  // compatible transform (i.e. "__esModule" has not been set), then set
-  // "default" to the CommonJS "module.exports" for node compatibility.
-  isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target,
-  mod
-));
-
-// ../lambda-microvm-runtime/dist/runtime.js
-var require_runtime = __commonJS({
-  "../lambda-microvm-runtime/dist/runtime.js"(exports2) {
-    "use strict";
-    Object.defineProperty(exports2, "__esModule", { value: true });
-    exports2.HOOK_SIGTERM_SECONDS = exports2.HOOK_PATHS = exports2.HOOK_NAMES = exports2.HOOK_PATH_PREFIX = void 0;
-    exports2.parseRunEnvelope = parseRunEnvelope;
-    exports2.createMicrovmRuntime = createMicrovmRuntime;
-    var node_http_1 = require("node:http");
-    exports2.HOOK_PATH_PREFIX = "/aws/lambda-microvms/runtime/v1";
-    exports2.HOOK_NAMES = ["ready", "validate", "run", "resume", "suspend", "terminate"];
-    exports2.HOOK_PATHS = Object.fromEntries(exports2.HOOK_NAMES.map((name) => [name, `${exports2.HOOK_PATH_PREFIX}/${name}`]));
-    var RUN_HOOK_PAYLOAD_MAX_BYTES = 16384;
-    var MAX_BODY_BYTES = RUN_HOOK_PAYLOAD_MAX_BYTES * 8;
-    exports2.HOOK_SIGTERM_SECONDS = 55;
-    var DEFAULT_TIMEOUT_SECONDS = exports2.HOOK_SIGTERM_SECONDS;
-    var DEFAULT_KILL_GRACE_SECONDS = 10;
-    function readBody(request) {
-      return new Promise((resolve, reject) => {
-        const chunks = [];
-        let size = 0;
-        let oversize = false;
-        request.on("data", (chunk) => {
-          if (oversize) {
-            return;
-          }
-          size += chunk.length;
-          if (size > MAX_BODY_BYTES) {
-            oversize = true;
-            resolve(null);
-            return;
-          }
-          chunks.push(chunk);
-        });
-        request.on("end", () => {
-          if (!oversize) {
-            resolve(Buffer.concat(chunks).toString("utf8"));
-          }
-        });
-        request.on("error", reject);
-      });
-    }
-    function parseRunEnvelope(rawBody) {
-      let parsed;
-      try {
-        parsed = JSON.parse(rawBody);
-      } catch {
-        throw new Error("run hook body is not valid JSON");
-      }
-      if (typeof parsed !== "object" || parsed === null || Array.isArray(parsed)) {
-        throw new Error("run hook body is not a JSON object");
-      }
-      const envelope = parsed;
-      const microvmId = envelope.microvmId;
-      const runHookPayload = envelope.runHookPayload;
-      if (typeof microvmId !== "string" || microvmId === "") {
-        throw new Error("run hook body is missing microvmId");
-      }
-      if (typeof runHookPayload !== "string") {
-        throw new Error("run hook body is missing runHookPayload");
-      }
-      return { microvmId, runHookPayload };
-    }
-    function createMicrovmRuntime({ hooksDir, spawn, hasScript, timeoutSeconds = DEFAULT_TIMEOUT_SECONDS, killGraceSeconds = DEFAULT_KILL_GRACE_SECONDS }) {
-      function runScript(hook, script, env, payload) {
-        return new Promise((resolve) => {
-          const child = spawn(script, { env });
-          let settled = false;
-          let escalation;
-          const settle = (outcome) => {
-            if (settled) {
-              return;
-            }
-            settled = true;
-            clearTimeout(timer);
-            resolve(outcome);
-          };
-          const timer = setTimeout(() => {
-            console.error(`[hook] hook=${hook} status=timeout limit_seconds=${timeoutSeconds}`);
-            child.kill("SIGTERM");
-            escalation = setTimeout(() => {
-              console.error(`[hook] hook=${hook} status=kill_escalated grace_seconds=${killGraceSeconds}`);
-              child.kill("SIGKILL");
-            }, killGraceSeconds * 1e3);
-            escalation.unref();
-            settle({ statusCode: 500, message: "hook timed out" });
-          }, timeoutSeconds * 1e3);
-          timer.unref();
-          child.once("error", (error) => {
-            console.error(`[hook] hook=${hook} status=spawn_failed error=${error.message}`);
-            settle({ statusCode: 500, message: "hook could not be started" });
-          });
-          child.once("exit", (code) => {
-            clearTimeout(escalation);
-            if (code === 0) {
-              settle({ statusCode: 200, message: hook });
-              return;
-            }
-            console.error(`[hook] hook=${hook} status=failed exit_code=${code}`);
-            settle({ statusCode: 500, message: "hook failed" });
-          });
-          if (payload !== void 0 && child.stdin) {
-            child.stdin.write(payload);
-            child.stdin.end();
-          } else {
-            child.stdin?.end();
-          }
-        });
-      }
-      async function dispatch(hook, rawBody) {
-        const env = { ...process.env };
-        let payload;
-        if (hook === "run") {
-          const envelope = parseRunEnvelope(rawBody);
-          env.MICROVM_ID = envelope.microvmId;
-          payload = envelope.runHookPayload;
-        }
-        const script = `${hooksDir}/${hook}`;
-        if (!hasScript(script)) {
-          return { statusCode: 200, message: `${hook} (no script)` };
-        }
-        const outcome = await runScript(hook, script, env, payload);
-        if (hook === "terminate" && outcome.statusCode !== 200) {
-          console.error("[hook] hook=terminate status=reported_ok_despite_failure");
-          return { statusCode: 200, message: "terminate (best effort)" };
-        }
-        return outcome;
-      }
-      return (0, node_http_1.createServer)((request, response) => {
-        const send = (statusCode, message) => {
-          response.writeHead(statusCode, { "content-type": "application/json" });
-          response.end(JSON.stringify({ message }));
-        };
-        const path = (request.url ?? "").split("?")[0];
-        const hook = exports2.HOOK_NAMES.find((name) => exports2.HOOK_PATHS[name] === path);
-        if (!hook) {
-          request.resume();
-          send(404, "unknown hook path");
-          return;
-        }
-        if (request.method !== "POST") {
-          request.resume();
-          send(405, "hook requires POST");
-          return;
-        }
-        void readBody(request).then(async (body) => {
-          if (body === null) {
-            send(413, "hook body too large");
-            return;
-          }
-          try {
-            const outcome = await dispatch(hook, body);
-            send(outcome.statusCode, outcome.message);
-          } catch (error) {
-            send(400, error instanceof Error ? error.message : "bad request");
-          }
-        }).catch((error) => {
-          console.error(`[hook] hook=${hook} status=unhandled_error error=${error instanceof Error ? error.message : String(error)}`);
-          send(500, "hook failed");
-        });
-      });
-    }
-  }
-});
-
-// ../lambda-microvm-runtime/dist/node-runtime.js
-var require_node_runtime = __commonJS({
-  "../lambda-microvm-runtime/dist/node-runtime.js"(exports2) {
-    "use strict";
-    Object.defineProperty(exports2, "__esModule", { value: true });
-    exports2.createNodeRuntime = createNodeRuntime2;
-    var node_child_process_1 = require("node:child_process");
-    var node_fs_1 = require("node:fs");
-    var runtime_1 = require_runtime();
-    function createNodeRuntime2({ hooksDir, timeoutSeconds, killGraceSeconds }) {
-      return (0, runtime_1.createMicrovmRuntime)({
-        hooksDir,
-        timeoutSeconds,
-        killGraceSeconds,
-        hasScript: (script) => {
-          try {
-            (0, node_fs_1.accessSync)(script, node_fs_1.constants.X_OK);
-            return true;
-          } catch {
-            return false;
-          }
-        },
-        spawn: (script, options) => {
-          const child = (0, node_child_process_1.spawn)(script, {
-            env: options.env,
-            // stdin is a pipe so the run payload never has to travel in argv.
-            stdio: ["pipe", "inherit", "inherit"],
-            // Its own process group, so a signal can reach the tree the script
-            // built — a hook shells out, and stopping only the shell orphans the
-            // synchroniser or probe it was waiting on.
-            detached: true
-          });
-          child.kill = (signal) => {
-            try {
-              process.kill(-child.pid, signal);
-              return true;
-            } catch (error) {
-              if (error.code !== "ESRCH") {
-                console.error(`[hook] script=${script} status=signal_failed signal=${String(signal)} error=${error instanceof Error ? error.message : String(error)}`);
-              }
-              return false;
-            }
-          };
-          child.stdin?.on("error", (error) => {
-            console.error(`[hook] script=${script} status=stdin_write_failed error=${error.message}`);
-          });
-          return child;
-        }
-      });
-    }
-  }
-});
-
-// ../lambda-microvm-runtime/dist/index.js
-var require_dist = __commonJS({
-  "../lambda-microvm-runtime/dist/index.js"(exports2) {
-    "use strict";
-    Object.defineProperty(exports2, "__esModule", { value: true });
-    exports2.createNodeRuntime = exports2.HOOK_SIGTERM_SECONDS = exports2.HOOK_PATH_PREFIX = exports2.HOOK_PATHS = exports2.HOOK_NAMES = exports2.parseRunEnvelope = exports2.createMicrovmRuntime = void 0;
-    var runtime_1 = require_runtime();
-    Object.defineProperty(exports2, "createMicrovmRuntime", { enumerable: true, get: function() {
-      return runtime_1.createMicrovmRuntime;
-    } });
-    Object.defineProperty(exports2, "parseRunEnvelope", { enumerable: true, get: function() {
-      return runtime_1.parseRunEnvelope;
-    } });
-    Object.defineProperty(exports2, "HOOK_NAMES", { enumerable: true, get: function() {
-      return runtime_1.HOOK_NAMES;
-    } });
-    Object.defineProperty(exports2, "HOOK_PATHS", { enumerable: true, get: function() {
-      return runtime_1.HOOK_PATHS;
-    } });
-    Object.defineProperty(exports2, "HOOK_PATH_PREFIX", { enumerable: true, get: function() {
-      return runtime_1.HOOK_PATH_PREFIX;
-    } });
-    Object.defineProperty(exports2, "HOOK_SIGTERM_SECONDS", { enumerable: true, get: function() {
-      return runtime_1.HOOK_SIGTERM_SECONDS;
-    } });
-    var node_runtime_1 = require_node_runtime();
-    Object.defineProperty(exports2, "createNodeRuntime", { enumerable: true, get: function() {
-      return node_runtime_1.createNodeRuntime;
-    } });
-  }
-});
-
-// ../../runner/docker-images/microvm/hook-server.ts
-var import_lambda_microvm_runtime = __toESM(require_dist());
-var HOOKS_DIR = "/etc/evident/hooks";
-var HOOKS_PORT = 8080;
-var server = (0, import_lambda_microvm_runtime.createNodeRuntime)({ hooksDir: HOOKS_DIR });
-var port = Number(process.env.HOOKS_PORT ?? HOOKS_PORT);
-server.listen(port, "0.0.0.0", () => {
-  console.log(`[hook] server listening on 0.0.0.0:${port}`);
-});