Files
PaperClipAI/tests/runner-e2e/daytona-image-content.ts
Dotta af3023f1e3 fix(runner): repair paid provider startup paths (#12769)
## Thinking Path

> - Paperclip manages AI agents that perform work.
> - Paperclip Runner connects durable task runs to local provider
processes.
> - The full-stack paid matrix exposed failures after the runner
integrity repair.
> - Verified JavaScript entrypoints lost their relative module graph
when Linux executed them through descriptor paths.
> - Returned provider startup errors also remained pending and became
indeterminate after recovery.
> - Sparse Codex tool lifecycle events lost the `write_document`
identity before task transcript projection.
> - This pull request repairs those three boundaries and makes the
structured-question fixture deterministic.
> - The benefit is repeatable provider startup, exact failure replay,
and correct inline Plan placement.

## Linked Issues or Issue Description

Refs #12721 and #12700.

**What happened?**

The paid runner matrix failed ACPX and OpenCode startup before provider
session creation. The runner journal then replaced the original startup
error with an indeterminate recovery result. Native Codex saved a Plan
but rendered it only as a fallback card. A legacy Claude waiting reply
could also echo the reserved terminal marker before the answer arrived.

**Expected behavior**

Verified JavaScript providers must start from immutable
descriptor-backed artifacts. Returned startup failures must persist as
terminal failed command results. Native tool lifecycle updates must
preserve the `write_document` boundary. Pre-answer fixture output must
not contain the reserved terminal marker.

**Steps to reproduce**

1. Run the local provider cells in the Runner Full-Stack E2E workflow.
2. Observe ACPX and OpenCode fail during `session.open` before provider
execution.
3. Observe recovery report `execution_indeterminate` instead of the
original startup error.
4. Run the native Codex Plan cell and observe the fallback Plan card
after the tool activity row.
5. Run the legacy Claude structured-question resume cell and observe an
early marker echo in waiting prose.

**Paperclip version or commit**

`0f9452101740835ce0b1488a204bf48acd5bafc3`

**Deployment mode**

Local development with the paid GitHub Actions acceptance workflow.

## What Changed

- Bundle the ACPX sidecar and OpenCode proxy as self-contained Node ESM
entrypoints before hashing and verified descriptor launch.
- Anchor ACPX dynamic provider package resolution at a
controller-derived provider-pack root and keep that root out of the
provider child environment.
- Persist executor-returned startup errors as redacted durable failed
command results while retaining indeterminate recovery for true process
death.
- Coalesce sparse native tool items by stable ID so a late
`write_document` name, input, and result reach the transcript boundary
once.
- Forbid the structured-question fixture from spelling or announcing its
reserved terminal marker before the user answers.

## Verification

- Rust and TypeScript regression tests cover durable failed replay, true
crash ambiguity, bundle closure, package-root derivation, environment
filtering, exact Codex tool lifecycle coalescing, and prompt
determinism.
- Local execution is intentionally limited to formatters and static diff
checks. GitHub Actions will run tests, type checks, builds, and security
checks.
- After ordinary CI is green, scoped paid cells will validate one ACPX
launch, one OpenCode launch, native Codex Plan projection, and legacy
Claude structured resume before a complete matrix rerun.
- Prior failing matrix:
https://github.com/paperclipai/paperclip/actions/runs/33682434315

## Risks

- Bundling changes the bytes covered by provider launch hashes.
Provider-pack generation already hashes the final built files.
- ACPX still loads qualified provider packages dynamically. The
controller supplies a normalized package root, while existing version,
digest, path, and descriptor checks remain active.
- Durable `failed` is terminal. Replays return the same redacted result
and do not execute the provider effect twice.

> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.

## Model Used

OpenAI Codex based on GPT-5 with agentic reasoning, repository
inspection, code editing, Git, parallel subagents, and GitHub Actions
coordination. The exact deployed snapshot and context-window size are
not exposed to this task.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either linked related public work or described the bug in
this PR
- [x] I have not referenced internal or instance-local Paperclip issues
or links
- [x] My branch name describes the change and contains no internal
ticket id
- [ ] I have run tests locally and they pass (intentionally deferred to
GitHub Actions)
- [x] I have added or updated tests where applicable
- [x] No documentation change is required for this runtime repair
- [x] I have considered and documented the risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge
2026-09-04 07:58:44 -05:00

280 lines
8.8 KiB
TypeScript

import { createHash } from "node:crypto";
import { lstat, readdir, readFile, readlink } from "node:fs/promises";
import path from "node:path";
import { fileURLToPath, pathToFileURL } from "node:url";
const repositoryRoot = path.resolve(import.meta.dirname, "../..");
export const DAYTONA_IMAGE_CONTENT_SCHEMA =
"paperclip-daytona-runner-image-content/v4";
export const DAYTONA_IMAGE_PLATFORM = "linux/amd64";
export const DAYTONA_IMAGE_DOCKERFILE_PATH = "docker/daytona-runner/Dockerfile";
// This is the audited dependency closure of docker/daytona-runner/Dockerfile.
// Keep it conservative: a false positive only rebuilds the image, while a
// missing input could incorrectly reuse an incompatible paid-test image.
export const DAYTONA_IMAGE_INPUT_PATHS = [
".dockerignore",
".npmrc",
"docker/daytona-runner/Dockerfile",
"package.json",
"patches",
"pnpm-lock.yaml",
"pnpm-workspace.yaml",
"scripts/link-plugin-dev-sdk.mjs",
"tsconfig.base.json",
"packages/paperclip-eval-kernel",
"packages/paperclip-runner",
] as const;
const ignoredGeneratedDirectoryPaths = new Set([
"packages/paperclip-runner/dist",
"packages/paperclip-runner/runner/target",
]);
// These paths do not contribute to the release runnerd binary or the
// executable/digested provider-pack runtime payload. They are also excluded
// from the real Docker build context by .dockerignore. Keep the two lists in
// lockstep: if a future build starts consuming one of these inputs, Docker must
// fail instead of publishing bytes that the content identity did not hash.
const ignoredRunnerDevelopmentDirectoryPaths = new Set([
"packages/paperclip-runner/devtools",
"packages/paperclip-runner/docs",
"packages/paperclip-runner/examples",
"packages/paperclip-runner/test",
"packages/paperclip-runner/test-fixtures",
"packages/paperclip-runner/test-support",
]);
const runnerDocumentationFilePattern = /\.md$/;
const runnerTestFilePattern = /\.(?:spec|test)\.(?:[cm]?[jt]sx?)$/;
const runnerRustIntegrationTestPathPattern =
/^packages\/paperclip-runner\/runner\/crates\/[^/]+\/tests(?:\/|$)/;
const runnerSmokeScriptPattern =
/^packages\/paperclip-runner\/scripts\/[^/]+-smoke\.mjs$/;
export interface DaytonaImageContentOptions {
repositoryRoot?: string;
inputPaths?: readonly string[];
platform?: string;
baseImages?: readonly string[];
frontendDigest?: string;
}
function compareNames(left: string, right: string): number {
return left < right ? -1 : left > right ? 1 : 0;
}
function normalizedRelativePath(value: string): string {
return value.split(path.sep).join("/");
}
function shouldIgnore(relativePath: string): boolean {
if (ignoredGeneratedDirectoryPaths.has(relativePath)) return true;
return relativePath.split("/").includes("node_modules");
}
function shouldIgnoreRunnerDevelopmentInput(relativePath: string): boolean {
if (!relativePath.startsWith("packages/paperclip-runner/")) return false;
if (ignoredRunnerDevelopmentDirectoryPaths.has(relativePath)) return true;
return (
runnerDocumentationFilePattern.test(relativePath) ||
runnerTestFilePattern.test(relativePath) ||
runnerRustIntegrationTestPathPattern.test(relativePath) ||
runnerSmokeScriptPattern.test(relativePath)
);
}
function updateRecord(
hash: ReturnType<typeof createHash>,
kind: string,
relativePath: string,
payload = "",
): void {
hash.update(kind);
hash.update("\0");
hash.update(relativePath);
hash.update("\0");
hash.update(payload);
hash.update("\0");
}
function assertPinnedBaseImage(reference: string): void {
if (!/@sha256:[0-9a-f]{64}$/.test(reference)) {
throw new Error(
`Daytona image base must use an immutable sha256 digest: ${reference}`,
);
}
}
function assertPinnedFrontendDigest(digest: string): void {
if (!/^sha256:[0-9a-f]{64}$/.test(digest)) {
throw new Error(
`Daytona Dockerfile frontend must use an immutable sha256 digest: ${digest}`,
);
}
}
export function extractDaytonaDockerfileFrontendDigest(
dockerfile: string,
): string {
const firstLine = dockerfile.split(/\r?\n/, 1)[0] ?? "";
const match = /^#\s*syntax=\S+@(sha256:[0-9a-f]{64})\s*$/.exec(firstLine);
if (!match) {
throw new Error(
"Daytona Dockerfile first line must pin its syntax frontend to an immutable sha256 digest",
);
}
const digest = match[1]!;
assertPinnedFrontendDigest(digest);
return digest;
}
export function extractDaytonaBaseImages(dockerfile: string): string[] {
const baseImages: string[] = [];
const stageAliases = new Set<string>();
for (const line of dockerfile.split(/\r?\n/)) {
if (!/^\s*FROM\s/i.test(line)) continue;
const match = /^\s*FROM(?:\s+--\S+)*\s+(\S+)(?:\s+AS\s+(\S+))?\s*$/i.exec(
line,
);
if (!match) {
throw new Error(`Cannot parse Daytona Dockerfile FROM line: ${line}`);
}
const reference = match[1]!;
if (!stageAliases.has(reference)) {
assertPinnedBaseImage(reference);
baseImages.push(reference);
}
if (match[2]) stageAliases.add(match[2]);
}
if (baseImages.length === 0) {
throw new Error("Daytona Dockerfile does not declare a base image");
}
return baseImages;
}
async function resolveBaseImages(
root: string,
explicitBaseImages?: readonly string[],
): Promise<string[]> {
const baseImages = explicitBaseImages
? [...explicitBaseImages]
: extractDaytonaBaseImages(
await readFile(path.join(root, DAYTONA_IMAGE_DOCKERFILE_PATH), "utf8"),
);
if (baseImages.length === 0) {
throw new Error("Daytona image content identity requires a base image");
}
for (const reference of baseImages) assertPinnedBaseImage(reference);
return baseImages.sort(compareNames);
}
async function resolveFrontendDigest(
root: string,
explicitFrontendDigest?: string,
): Promise<string> {
const digest =
explicitFrontendDigest ??
extractDaytonaDockerfileFrontendDigest(
await readFile(path.join(root, DAYTONA_IMAGE_DOCKERFILE_PATH), "utf8"),
);
assertPinnedFrontendDigest(digest);
return digest;
}
async function hashEntry(
hash: ReturnType<typeof createHash>,
root: string,
relativePath: string,
): Promise<void> {
const normalizedPath = normalizedRelativePath(relativePath);
if (
shouldIgnore(normalizedPath) ||
shouldIgnoreRunnerDevelopmentInput(normalizedPath)
) {
return;
}
const absolutePath = path.resolve(root, relativePath);
const relativeFromRoot = path.relative(root, absolutePath);
if (
relativeFromRoot.startsWith(`..${path.sep}`) ||
relativeFromRoot === ".." ||
path.isAbsolute(relativeFromRoot)
) {
throw new Error(
`Daytona image input escapes repository root: ${relativePath}`,
);
}
const stats = await lstat(absolutePath);
if (stats.isDirectory()) {
updateRecord(hash, "directory", normalizedPath);
const entries = await readdir(absolutePath, { withFileTypes: true });
entries.sort((left, right) => compareNames(left.name, right.name));
for (const entry of entries) {
await hashEntry(hash, root, path.join(relativePath, entry.name));
}
return;
}
if (stats.isSymbolicLink()) {
updateRecord(hash, "symlink", normalizedPath, await readlink(absolutePath));
return;
}
if (stats.isFile()) {
const executable =
(stats.mode & 0o111) === 0 ? "non-executable" : "executable";
updateRecord(hash, "file", normalizedPath, executable);
hash.update(await readFile(absolutePath));
hash.update("\0");
return;
}
throw new Error(`Unsupported Daytona image input type: ${relativePath}`);
}
export async function computeDaytonaImageContentId(
options: DaytonaImageContentOptions = {},
): Promise<string> {
const root = path.resolve(options.repositoryRoot ?? repositoryRoot);
const inputPaths = [
...(options.inputPaths ?? DAYTONA_IMAGE_INPUT_PATHS),
].sort(compareNames);
const hash = createHash("sha256");
updateRecord(
hash,
"contract",
DAYTONA_IMAGE_CONTENT_SCHEMA,
options.platform ?? DAYTONA_IMAGE_PLATFORM,
);
updateRecord(
hash,
"dockerfile-frontend",
await resolveFrontendDigest(root, options.frontendDigest),
);
for (const baseImage of await resolveBaseImages(root, options.baseImages)) {
updateRecord(hash, "base-image", baseImage);
}
for (const inputPath of inputPaths) {
await hashEntry(hash, root, inputPath);
}
return hash.digest("hex");
}
const invokedPath = process.argv[1] ? path.resolve(process.argv[1]) : null;
if (invokedPath && import.meta.url === pathToFileURL(invokedPath).href) {
computeDaytonaImageContentId()
.then((contentId) => process.stdout.write(`${contentId}\n`))
.catch((error: unknown) => {
process.stderr.write(
`${error instanceof Error ? error.message : String(error)}\n`,
);
process.exitCode = 1;
});
}
export const DAYTONA_IMAGE_CONTENT_SCRIPT = fileURLToPath(import.meta.url);