Files
PaperClipAI/docker/daytona-runner
cab4263dc9 feat(claude-local): add Sonnet 5.5 and refresh the qualified Claude runtime (#14993)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work.
> - The Claude local adapter lists models for users with a Claude
subscription.
> - The list needs Claude Sonnet 5.5 and its supported effort levels.
> - Sonnet 5.5 needs Claude Code 2.1.284 or later on both execution
paths.
> - The qualified ACP runtime previously used Claude Code 2.1.280.
> - This change adds Sonnet 5.5 and pins Agent SDK 0.3.286, which
includes Claude Code 2.1.286.
> - Users can select the model and run it with a qualified runtime.

## Linked Issues or Issue Description

Refs #3936. This replaces #14816 because the maintainer integration
cannot write to the contributor fork.

Thank you to @SkilLab-Tech for the model support, runtime refresh,
tests, and platform digest verification. This branch preserves both
original commits: `8d2f3261af61a2ac1120e51e8a8618732ace543b` and
`e68d1d002a3ed745f016fb11c50ac5a3c5a9ff8d`.

Related work:

- #14917 added Claude model ordering. This branch includes that merged
change and resolves its conflicts with #14816.
- #14942 updates the other models and harnesses. It remains separate.
Its matching Sonnet effort and CLI-gate changes are identical. Both PRs
merge with master. The second PR will need a rebase after the first
merges because adjacent runtime-pin and test edits conflict.
- #14954 is another Sonnet 5.5 change. It overlaps with the model
additions but does not include the qualified runtime refresh.
- #14039 makes the per-task effort picker model-aware. #3937 is also
related to effort selection.

The original author checked the [Claude Code
changelog](https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md)
and [effort
documentation](https://platform.claude.com/docs/en/build-with-claude/effort)
on 2026-10-01.

## What Changed

- Add the direct `claude-sonnet-5-5` model and Low, Medium, High,
X-High, and Max effort levels.
- Require Claude Code 2.1.284 or later for that model on the CLI path.
- Put Sonnet 5.5 after Opus 5.5 in the current-model group. Keep Sonnet
5 in the older-model group.
- Retain the Sonnet 5.5 assertions and the model-order assertions in the
server tests.
- Pin Agent SDK 0.3.286 and Claude Code 2.1.286 across overrides,
integrity digests, qualified profiles, Rust provider pins, and the
Daytona version check.
- Update the related adapter and runtime documentation.

## Verification

Local verification uses the resolved source tree and pnpm 9.15.4. Model
tests passed on Node 25.9.0. Runtime integrity tests use CI's Node
24.21.0.

- Five focused Claude test files pass: 62 tests. They cover model
defaults, model ordering, CLI gates, and remote execution probes.
- Server model-list and UI setup tests pass: 30 tests.
- The Claude adapter typecheck passes.
- Runner integrity and qualification tests pass on Node 24.21.0: 78
tests. Four descriptor-loader tests fail on Node 25.9.0; all four pass
on the CI version.
- The runner package contract passes: 10 tests.
- Full local typecheck stopped with exit 137 in the database package
under the container's 4 GB memory limit. The production build reached
the runner Rust build, then stopped because `cargo` is absent.
- The full stable local Vitest run was stopped after all current-head CI
test shards passed. It did not complete locally. The 180 focused tests
listed above passed.
- `git diff --check` passes. The branch changes 20 files against master.
It has no lockfile or workflow changes.
- The original author verified all three platform digests against
registry integrity and ran the Linux executable. Its version was
`2.1.286 (Claude Code)`. See #14816 for that evidence.
- Greptile reviewed head `6db3d3f1` and gave 5/5 with zero comments.
Both Superagent scans and Commitperclip pass. All current-head CI jobs
pass, including build, typecheck, Rust, test shards, browser tests, and
the canary dry run.

## Risks

- The controller and provider pack must use matching runtime pins.
Deploy them together.
- CI owns `pnpm-lock.yaml`. The master lockfile refresh must resolve the
SDK override. Refresh the Daytona lock digest with that lockfile.
- Images built with Claude Code older than 2.1.284 need a rebuild before
the CLI path can use Sonnet 5.5.
- The runtime remains at SDK 0.3.286. This PR does not take the later
0.3.287 patch.
- A live Sonnet 5.5 session and a Daytona image build are not part of
the local verification.

## Model Used

- Original work: Anthropic Claude Code, `claude-sonnet-5-5`. Review:
`claude-opus-5-5`. The author reported `xhigh` effort, tool use, and
code execution. The original context window was not reported.
- Merge repair and PR preparation: OpenAI Codex, based on GPT-6, with
tool use and code execution. The runtime does not expose the exact model
identifier or context window in this session.

## 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 (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

## Squash Attribution

Keep these trailers in the squash commit to preserve the original author
and AI attribution:

```text
Co-Authored-By: Claude Code (Ivan) <SkilLab-Tech@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Paperclip <noreply@paperclip.ing>
```

---------

Co-authored-by: Claude Code (Ivan) <ivan@skillab.com.br>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-10-05 11:03:20 -07:00
..

Paperclip Daytona runner image

This image is the Paperclip Cloud fleet sandbox image plus a source-built paperclip-runnerd and immutable provider pack. The pack contains Node 24.11, OpenCode 1.18.32, the compiled OpenCode proxy, ACPX 0.13.1 sidecar, qualified ACP agents, and the production lockfile. Its manifest digests each executable bridge and binds the pack to the runner source revision, avoiding artifact upload and npm installation on every fresh lease.

The fleet pins are intentionally copied from paperclip-cloud/fleet-sandbox-image/Dockerfile. Update both definitions together until the fleet base is published as a stable image that this Dockerfile can extend directly.

Harness versions

The October 1, 2026 refresh pins Codex 0.156.0, Claude Agent SDK 0.3.286 (Claude Code 2.1.286), and OpenCode 1.18.32 in the shared provider pack. Claude Code 2.1.280 is the minimum for Opus 5.5 and 2.1.284 is the minimum for Sonnet 5.5; the pinned runtime also supports Fable 5.1. Codex uses the current GPT-6 Sol and Luna model IDs. Grok CLI 1.0.41 supports the current Grok 4.7 model family.

Keep the patched ACP bridge versions separate from their CLI runtime pins. Their executable digests do not change when only the runtime dependency changes. Refresh the runtime executable digests from integrity-verified npm release archives for every supported platform, and keep the native runner, provider manifest, and remote controller version checks aligned.

Build and verify

Run pnpm --filter @paperclipai/paperclip-runner test:opencode:qualification after installing dependencies to exercise the actual pinned OpenCode executable. It checks health/version, session creation and retrieval, SSE messages, an async prompt, and session deletion against a loopback mock provider. It uses an isolated home, starts no paid model request, and retires its process group. Set PAPERCLIP_TEST_OPENCODE_BINARY to the materialized Linux executable when qualifying an assembled provider pack.

The fleet image is currently amd64-only because the pinned Cursor and GitHub CLI checksums cover amd64.

content_id="$(pnpm --silent test:e2e:runner:image-id)"
docker buildx build \
  --platform linux/amd64 \
  --build-arg PAPERCLIP_RUNNER_CONTENT_ID="${content_id}" \
  --build-arg PAPERCLIP_RUNNER_SOURCE_REVISION="$(git rev-parse HEAD)" \
  --tag "paperclip-daytona-runner:e2e-content-${content_id}" \
  --load \
  --file docker/daytona-runner/Dockerfile \
  .

docker run --rm --platform linux/amd64 \
  --entrypoint paperclip-runnerd \
  "paperclip-daytona-runner:e2e-content-${content_id}" \
  --build-metadata

The metadata must advertise dial_ws_loopback, dial_wss, and listen_ws. The explicit entrypoint is needed only for this local probe because Daytona's base image uses its own long-running sandbox entrypoint.

test:e2e:runner:image-id hashes the audited Docker build dependency closure, target platform, the immutable Dockerfile syntax-frontend digest, and every immutable FROM reference. It fails before the paid workflow can build when the frontend or a base is not pinned to a sha256 digest. When updating the syntax version, resolve and review its registry digest and update both values in the first Dockerfile line. Git commits that do not change those inputs reuse the same content tag.

PAPERCLIP_RUNNER_SOURCE_REVISION remains the full Git SHA that built the first published copy and is retained as provenance rather than cache identity.

Use in Paperclip

Publish the image to a registry Daytona can pull, or use the environment editor's Configure image flow to produce a Daytona snapshot. Set the environment image to that immutable tag or snapshot. Paperclip probes the sandbox user's PATH for paperclip-runnerd and codex and checks /opt/paperclip-runner/provider-pack for OpenCode and ACPX. It uses the pack only when its complete manifest matches the controller's build-owned pack; otherwise it stages the pack configured by PAPERCLIP_RUNNER_REMOTE_PROVIDER_PACK_PATH. Remote OpenCode and ACPX never fall back to host-local processes.

Do not promote paperclip-runner-e2e-20260826-v2 for OpenCode or ACPX. Build a new immutable image or snapshot from a clean committed revision and pass that full Git SHA as PAPERCLIP_RUNNER_SOURCE_REVISION.

Do not bake provider credentials, Paperclip bootstrap tickets, or Daytona preview tokens into this image. They remain per-run secret material.

The provider-pack build pins the official Linux x64 Node 24.21.0 image by manifest digest. Its bundled Undici is 7.29.1, which fixes GHSA-3wwx-pv8p-q78v. Pi separately verifies its private Node executable and nested npm dependency; changing the outer interpreter does not replace either provider-owned pin.

Provider CLI updates are manifest-only changes: repository CI owns the root lockfile. Resolve the complete workspace manifest graph in the build context before invoking Docker, matching CI when a source commit precedes the lockfile bot. The trusted workflow supplies this resolved lockfile as an immutable artifact. The complete resolved lockfile must match PAPERCLIP_RUNNER_LOCK_SHA256 before package installation or lifecycle execution. Review and refresh that digest with source dependency changes; registry-time resolution drift fails closed. The Product E2E workflow resolves one lockfile before the image build. It verifies the downloaded artifact, then passes that artifact's SHA-256 as the PAPERCLIP_RUNNER_LOCK_SHA256 build argument. The Dockerfile checks the resolved lock against this value before installation. The fixed Dockerfile default is for standalone builds; it must not replace a campaign's verified lock digest. Refresh the default from the clean tracked lockfile using the exact pnpm install --resolution-only --ignore-scripts --no-frozen-lockfile command, and verify a second resolution preserves the digest. A lockfile left by a filtered or incremental install can retain stale importer patch identities. Refresh exact runtime versions and qualification digests together; do not download dependencies when a task starts.

Candidate ACP qualification assets

Provider branches can build their pinned assets with node packages/paperclip-runner/scripts/build-provider-pack.mjs /absolute/pack --candidate-providers=cursor (or copilot or pi). The source revision must include the named provider's builder. Assets are installed at build time under provider-assets/<provider>/<platform>-<architecture>. The pack manifest binds each complete asset tree. Runtime admission separately checks the provider's source-owned closure pins and copies a verified launch snapshot. A pack with candidate assets does not qualify or enable that provider.

For an isolated Linux x64 Daytona qualification image, pass --build-arg PAPERCLIP_RUNNER_CANDIDATE_PROVIDERS=cursor with the normal build arguments. Compute its content ID with the same selection: pnpm --silent test:e2e:runner:image-id --candidate-providers=cursor. Candidate assets and the default pack have distinct image identities. Never reuse the default image's content ID for a candidate build. Use each provider branch's recorded version and explicit model. Keep credentials out of images. Paid qualification requires bound provider and Daytona credentials, inspectable spend, and the shared $100 ceiling recorded in the capability report.

The pack builder tests its copied Node interpreter after relocation. Use a standalone Node distribution if the host interpreter depends on a package manager's private shared libraries. Pi additionally pins its complete portable interpreter and npm dependency graph. Refresh exact runtime versions and qualification digests together; never download dependencies when a task starts. Grok's additive native ACP profile keeps its qualified 1.0.13 executable at the verified sandbox prerequisite path. It does not replace the legacy adapter's grok command on PATH.

Native Grok is an image prerequisite at /opt/paperclip/providers/grok/1.0.13/grok. Its checksum-verified provisioning is separate from the provider pack, which ships only the built-in launcher. Public npm installation never downloads this binary.