Commit Graph
4 Commits
Author SHA1 Message Date
Nicky LeachandClaude Fable 5.2 647641a99b feat(ci): let docker.yml copy the published Runner image, behind a repository variable (#21340)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - The Runner is the part that executes agent work. `master` publishes
it alone as an attested image, once for each set of its inputs
> - Since #21332 the application image takes its Runner only from the
`runner-artifacts` stage, and a build can replace that stage with the
published image
> - No workflow does that yet, so every application image still builds
its own Runner, and nothing on an application image says which Runner it
has
> - A build that copies the published Runner must also build the
lockfile that Runner was built from. The workflow refreshes the lockfile
before every build, and the server build then refuses the published
Runner
> - This pull request lets `docker.yml` copy the published Runner image
and label the application image with it, behind a repository variable
that is off, and makes that build use the committed lockfile
> - The benefit is one Runner for all images of the same inputs, named
on the image, with a switch that goes back to today's builds without a
revert

## Linked Issues or Issue Description

Refs #21204, #21229, #21230 (the Runner image), #21306 (the resolver),
#21307 (the report-only job), #21308 (the server reads a pack from its
Runner image), and #21331, #21332, #21333 (the application image takes
its Runner from `runner-artifacts`). No issue exists for this change.

**Problem or motivation**

`docker.yml` builds the application image on `master` and on tags. Each
build makes its own Runner. The report-only job that runs the resolver
on `master` feeds no build. Downstream image composers cannot tell from
an application image which Runner it has.

There is a second problem, which only shows when a build copies the
published Runner. The Runner's build stamp covers the whole
`pnpm-lock.yaml`. The step that vendors the Runner into the server fails
when the stamp does not match the tree. `docker.yml` refreshes
`pnpm-lock.yaml` on the runner before every build. The refresh rewrites
the file whenever the manifest of any package is ahead of it, also a
package the Runner does not read. On the tree of this pull request it
rewrites the file even for a pristine checkout: the registry changed the
deprecation text of one package. The published Runner was built from the
committed file, so the image build would fail.

**Proposed solution**

One job resolves the Runner once for the run. Both architecture builds
need it and add its answer to their build: a build context that replaces
the `runner-artifacts` stage, two labels, and the build argument with
which the build checks the Runner. A build that copies the published
Runner does not refresh the lockfile: it builds the committed tree, the
tree the Runner was published for. The job runs only when the repository
variable `RUNNER_IMAGE_CONSUME_ENABLED` is exactly `true`. This pull
request merges with the variable unset, so no image changes when it
merges.

**Alternatives considered**

- Resolve in each architecture build. The two builds could then
disagree, and one image index would hold two Runners.
- Write the labels in the `Dockerfile`. A `LABEL` cannot be left out for
an inline build.
- For the lockfile, copy the Runner image's `/pnpm-lock.yaml` over
`/app/pnpm-lock.yaml` in the `build` stage. Nothing after that step
reads the file: the two vendor checks are its only readers in `build`,
`production` copies it and takes `node_modules` from `deps`, both
`cloud-*` stages install with `--ignore-workspace --no-lockfile`, and
the server reads only a plugin's own lockfile and the pack's own
lockfile at run time. But the image would then ship a lockfile that its
`node_modules` did not come from, the copy changes that file's layer for
every inline build, and each `Dockerfile` edit makes a new Runner image.
- For the lockfile, bind-mount the Runner image's lockfile over the file
for the two steps that run the vendor check. This also edits the
`Dockerfile`, and it makes the lockfile part of the check always true
for a published Runner.
- For the lockfile, let the resolver refresh and compare. The answer
would depend on the registry at two different moments.
- Skipping the refresh needs no `Dockerfile` edit, keeps the stamp check
as strict as it is, and builds exactly the tree the inputs digest was
computed from.

**Roadmap alignment**

This is build infrastructure. It adds no product feature and does not
overlap an item in `ROADMAP.md`.

This pull request is built on `runner/app-image-consume`. Merge in this
order:

1. #21331, #21332, #21333: the application image takes its Runner from
the `runner-artifacts` stage, checks the inputs digest a build names,
and stops early on a stale lockfile.
2. `runner/app-image-consume`: the wait and the `unsupported` answer of
the resolver.
3. This pull request.

The same change for the preview cloud image build in `release.yml` is
deferred. It has no pull request yet.

## What Changed

- `.github/workflows/docker.yml`: the job `runner_image` replaces
`runner_image_report`.
- It runs only when `vars.RUNNER_IMAGE_CONSUME_ENABLED == 'true'`, in
this repository. A fork never runs it.
- Steps: `inputs` with no credential; `resolve`; then `gh attestation
verify` of the full index against the signer `runner-image.yml` on
`refs/heads/master`.
- For a `master` push, `resolve` runs with `--wait` and the job token.
For a tag or a manual run it does not wait and receives no token.
- Only the verification step writes the job's outputs. An image that
does not verify fails that step with a warning and writes no output. The
job is `continue-on-error`: the job is red, the run is not, and the
builds are inline.
- Permissions: `contents: read`, `actions: read`, `attestations: read`.
No registry login, no write.
- Limits: 5 minutes for `inputs`, 55 for the lookup, 5 for the
verification, 70 for the job. The wait of 40 minutes ends inside them.
- `build-and-push` needs `runner_image` and adds its outputs to the
build: `build-contexts`, one more line of `build-args`
(`PAPERCLIP_RUNNER_EXPECT_INPUTS_DIGEST`), two more lines of `labels`.
With empty outputs these are the values the build had before.
- The step "Refresh lockfile for Docker build context" runs only when
the answer is not `published`.
- The step that looks for a merge-queue image of the commit runs only
when the answer is not `published`. The two steps after it already
depend on its result.
- Labels on an image that copied the published Runner:
`io.paperclip.runner.image=ghcr.io/paperclipai/paperclip-runner@sha256:<full
index digest>` and `io.paperclip.runner.inputs-digest=<digest>`. An
image without them made its own Runner.
- The `Dockerfile` does not change. The Runner image digest does not
move.
- Tests that pin lines this changes follow:
`scripts/merge-group-image.test.mjs` (the condition of the merge-queue
lookup), `scripts/reuse-canary-image.test.mjs` (the `needs` of the build
job), and `docker-build-stamp.test.ts` (no workflow names a build
context; `docker.yml` now names exactly one).
- `server/src/__tests__/docker-runner-workflows.test.ts`: the
report-only tests give way to tests of the new job.
- `server/scripts/prepare-runner-vendor.test.mjs`: one fixture test of
the lockfile case.
- `doc/runner-image-contract.md` and `doc/DOCKER.md`: how an application
image gets its Runner, the labels and what their absence means, the
variable, the wait, the verification, the lockfile rule, the merge-queue
rule and the rollback.

## Verification

### Real builds

`docker buildx build --target build --platform linux/arm64 --output
type=cacheonly`, with a `docker-container` builder made for this test.
The Runner image is the publication form
(`PAPERCLIP_RUNNER_IMAGE_PUBLICATION=1`, `--target runner-artifacts`)
built from a clean export of this branch, exported as an OCI layout:
inputs `c4f9d089…34f0`, image `sha256:bef7e9cb…ffef`. The application
builds receive `--build-context
runner-artifacts=oci-layout://…@sha256:bef7e9cb…` and
`PAPERCLIP_RUNNER_EXPECT_INPUTS_DIGEST=c4f9d089…`. The refresh command
is the one in the workflow, run with pnpm 9.15.4 in the export: `pnpm
install --resolution-only --ignore-scripts --no-frozen-lockfile`.

| Build context | Result |
| --- | --- |
| The committed tree | Passes in 78 s. Both vendor checks: `reusing the
current Runner build (inputs unchanged …; 1333 files and 1860 outputs
checked)` |
| `ui/package.json`: `clsx` `^2.1.1` → `^2.1.0`, then the refresh
command (what the workflow did before this change) | Fails at `RUN node
server/scripts/prepare-runner-vendor.mjs`:
`PAPERCLIP_RUNNER_VENDOR_REQUIRE_CURRENT=1 forbids the Runner rebuild
this step needs (inputs changed since the recorded build; 1333 files
checked)` |
| The committed tree, then only the refresh command | Fails the same
way. The refresh changed only the `deprecated` text of one package |
| The same `ui/package.json` change with no refresh (what the workflow
does now for a `published` answer) | Passes in 66 s. `deps` resolves the
manifest inside the image; both vendor checks reuse the Runner |
| One line added to `packages/paperclip-runner/src/index.ts` | Fails at
the vendor step, as before: a Runner of other inputs is still refused |

Lockfile behind the Runner's own manifests
(`packages/paperclip-runner/package.json`: `smol-toml` `^1.9.1` →
`^1.9.0`), no build context:

| Step | Result |
| --- | --- |
| `runner-image-resolve.mjs inputs`, then `resolve` | `lock: behind`, no
digest; `inline (lock-behind)` |
| Build without the refresh | Fails in 4 s at the `runner-lockfile`
stage: `ERROR: pnpm-lock.yaml is behind the manifests the Runner is
installed from …` |
| Build after the refresh command (what the workflow does for an
`inline` answer) | Passes in 193 s: the `runner-lockfile` stage passes,
the build makes its own Runner with its pack, and both vendor checks
reuse it |

### Tests

macOS (Node 26.12.0) and Linux (`node:24.22.3-bookworm`, linux/arm64,
with `jq` and `zstd` added), on this branch:

- `pnpm exec vitest run` on `docker-runner-workflows`,
`runner-image-resolve`, `runner-image-workflow`, `docker-build-stamp`,
`container-init-reaping`, `cloud-image-bundled-plugins`,
`docker-image-layer-reuse`, `cloud-image-sentry`,
`runner-lockfile-guards` and `runner-image-identity` in
`server/src/__tests__`, and
`server/scripts/prepare-runner-vendor.test.mjs`: 225 tests in 11 files
pass on Linux; on macOS 224 pass and 1 is skipped (it needs bash 4.4).
- `node --test .github/scripts/tests/*.test.mjs
scripts/merge-group-image.test.mjs scripts/reuse-canary-image.test.mjs
scripts/cloud-source-verification.test.mjs
scripts/standard-image-contract.test.mjs
scripts/preview-artifacts.test.mjs
scripts/__tests__/release-verify-workflow.test.mjs
scripts/__tests__/release-dry-run-notes.test.mjs`: 991 of 991 pass on
both.
- The changed test files typecheck with the options of
`server/src/__tests__/tsconfig.json`.

What the workflow tests hold:

- Variable off: the job's condition is the variable and the repository,
with no alternative. With empty outputs, the build step has exactly the
3 build arguments, 0 build contexts and 1 label entry it had before, and
the refresh step runs.
- Published: both architecture builds are one job that reads only the
outputs of `runner_image`, with no architecture in those lines.
- The verification step runs as real shell against a stub `gh`. When
`gh` succeeds it writes the context, the labels and the build argument.
When `gh` fails the step exits 1, writes no output and prints the
warning. The step has no `continue-on-error`; the job has. The condition
of the build job does not name the resolver job and has no `success()`.
- The lookup step runs as real shell against a stub `node`. It passes
`--wait` only when the event is a `master` push.
- Credentials: only the lookup and the verification steps have a token.
The steps that build have no `env`.
- The refresh step has the condition `needs.runner_image.outputs.mode !=
'published'`, and it is the only `pnpm install` of the build job.
- The fixture test gives the vendor step a Runner that another tree
built for the committed lockfile. The step reuses it. After a change of
one specifier of another package in the lockfile the step fails and
names the lockfile. With the committed lockfile back it reuses the
Runner again.
- The merge-queue workflow and the pull-request check do not name the
resolver, the variable, a build context or a Runner label.

Read-only checks against the real services:

- `gh attestation verify
oci://ghcr.io/paperclipai/paperclip-runner@sha256:0cd8250b…1bf7 --repo
paperclipai/paperclip --signer-workflow
paperclipai/paperclip/.github/workflows/runner-image.yml --source-ref
refs/heads/master`: exit 0 for the full index, and exit 0 for the slim
index `sha256:69ab2001…69b3`.
- The same command with `--signer-workflow …/docker.yml`: exit 1 for
both. With `--source-ref refs/heads/not-master`: exit 1. With `--repo
cli/cli`: exit 1. For a digest that does not exist: exit 1. With no
token: exit 4.
- The parsed YAML of `docker.yml` differs from its base only in the
resolver job, the `needs` of the build job, the conditions of the
refresh step and of the merge-queue lookup, and the three build inputs.

Never run:

- `docker.yml`. Only GitHub can run it, and it has no pull-request
trigger.
- A build with the variable on. The first one is the first run after a
maintainer sets the variable.
- A build that pulls the published Runner from the registry as
`docker-image://`. The real builds used the same image form from a local
OCI layout.
- An `amd64` build.
- The `production` and `cloud` targets. The real builds stopped at
`build`, the last stage that checks the Runner.

## Risks

- With the variable on, an application build on `master` starts after
the resolver job. That is 2 to 3 minutes when the image exists, and up
to 40 minutes from the push when the commit's own Runner publication is
in progress. `docker.yml` holds one concurrency slot for each ref, so a
long wait can make more `master` commits lose their image build to a
newer push.
- A build that copies the published Runner ships the committed
`pnpm-lock.yaml` in `/app`, while `deps` resolves newer manifests of
other packages inside the image. A local `docker build` has always done
this; a CI image did not. Nothing in the image reads that file.
- Any other build that gives the `Dockerfile` a published Runner and a
refreshed lockfile fails at the vendor step. That is the intended answer
to a tree the Runner was not built for, and the message names the
lockfile.
- The pack of an image that copied the published Runner names the commit
that published the Runner (`runnerSourceRevision`), which can be older
than the commit of the image. A consumer that compares the two must
accept that before the variable is set.
- A failed verification makes the resolver job red on a green run. That
is on purpose: it means a published tag whose attestation does not
verify.
- `scripts/merge-group-image.mjs` refuses to promote a merge-queue image
for a push that changes `docker.yml`. The push that lands this change is
such a push. The merge queue is not enabled.
- With the variable off there is no report of the resolver on `master`
any more: the job that replaced the report-only job does not run.
- Rollback: set the variable to another value or delete it. The next run
builds inline, with the refresh. Images that exist do not change.

> 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

- Claude Fable 5.2 (`claude-fable-5-2`), Anthropic, used through Claude
Code with tool use (shell, file edits, local Docker builds, read-only
calls to the GitHub API and the container registry).

## 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

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-authored-by: Claude Fable 5.2 <noreply@anthropic.com>
2026-11-13 14:47:20 -08:00
Nicky LeachandClaude Fable 5.2 f79e2c5d84 feat(ci): add the Runner image resolver and a no-rebuild mode for the Runner vendor step (#21306)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - The Runner is the part that executes agent work: a daemon
(`paperclip-runnerd`) and a provider pack of agent tools
> - `master` now publishes the Runner alone as the image
`ghcr.io/paperclipai/paperclip-runner`, one time for each set of Runner
inputs (#21204, #21229, #21230)
> - The application image still compiles its own Runner on every build.
The next steps make it copy the published Runner instead
> - Before a build can do that, it must answer one question the same way
as every other build: "does a published Runner exist for exactly this
tree?" It must also fail clearly when the Runner it was given does not
match the tree
> - This pull request adds the script that answers the question, and a
mode of the Runner vendor step that fails instead of rebuilding
> - The benefit is that the later change to the image build is small,
and that a wrong or unreachable answer can only mean "build the Runner
yourself"

## Linked Issues or Issue Description

Refs #21204, #21229 and #21230 (the Runner image). No issue exists for
this change.

**Problem or motivation**

Three builders make images that contain the Runner: `docker.yml`, the
merge-queue build and the release preview build. Each one would have to
repeat three steps to find the published Runner of its tree: check the
lockfile, compute the inputs digest in the build, and look up
`inputs-<digest>`. Three copies can decide differently. A build that
copies a Runner built elsewhere also has no Rust toolchain. If the
copied Runner does not match the tree, `prepare-runner-vendor.mjs`
starts a rebuild, and the build fails late at `cargo` with a message
that hides the cause.

**Proposed solution**

Add `scripts/runner-image-resolve.mjs`, the one place that decides. Add
`PAPERCLIP_RUNNER_VENDOR_REQUIRE_CURRENT=1` to
`prepare-runner-vendor.mjs`. Nothing calls the resolver and nothing sets
the variable in this pull request, so no workflow job changes.

**Alternatives considered**

- One command that builds and looks up. Two commands keep a registry
credential, when a caller has one, out of the step that builds.
- A wait for the publication of the commit. The decision whether the
application build waits is open, so the resolver has no wait mode. It
reports `absent` and stops.
- A build argument that names the expected inputs. No `Dockerfile`
declares one yet. The change that makes the application image copy the
Runner will add both ends together.

**Roadmap alignment**

This is build infrastructure. It adds no product feature and does not
overlap an item in `ROADMAP.md`.

This is part 1 of 4 of a stack ([stack
#21309](https://github.com/paperclipai/paperclip/stack/21309)). Merge
the parts in order:

1. #21306: the resolver and the no-rebuild mode (this pull request)
2. #21307: the report-only job in `docker.yml` and the pull-request
check of the application image's Runner path
3. #21308: reading a retained provider pack from its Runner image
4. #21310: a registry build cache for the Runner image build

## What Changed

- `scripts/runner-image-resolve.mjs` (new, 185 lines), with two commands
that share a directory:
- `inputs --dir <dir>` exports the `runner-artifacts-context` target,
runs the lockfile check on that export, and builds the
`runner-artifacts-identity` target when the lockfile is current. It
reads no credential. A lockfile that is behind stops there: no image is
published for such a tree.
- `resolve --dir <dir>` looks up `inputs-<digest>` and prints `mode`,
`reason`, `inputs_digest`, `image`, `slim_image`, `build_contexts` and
`labels`. It also writes them to `$GITHUB_OUTPUT` and a short report to
`$GITHUB_STEP_SUMMARY`.
- `mode` is `published` only for a complete publication whose full index
names these inputs and both platforms. Both images are pinned by index
digest. The slim digest is the one the full index names in
`io.paperclip.runner.slim-digest`; the slim tag is never read on its
own.
- Every other answer is `inline`, with the last four values empty:

  | `reason` | Meaning |
  | --- | --- |
  | `lock-behind` | `pnpm-lock.yaml` is behind the Runner's manifests |
  | `absent` | no tag exists |
| `incomplete`, `broken` | a publication stopped, or its two tags do not
belong together |
| `mismatch` | the full index names other inputs, or lacks a platform |
| `unknown` | the registry refused, failed or could not be reached, or
the lookup threw |

- The resolver reuses `runnerLockState`, `readRunnerImageManifest`,
`runnerPublicationState` and `runnerImageTags` from
`scripts/runner-image-publish.mjs`. That file now exports
`environmentCredentials`; it has no other change.
- `server/scripts/prepare-runner-vendor.mjs`: with
`PAPERCLIP_RUNNER_VENDOR_REQUIRE_CURRENT=1`, the step reuses a current
build as before. When the build is not current, it prints the reason,
starts no rebuild, leaves the stamp in place and exits with status 1. It
also fails when `PAPERCLIP_RUNNER_VENDOR_REBUILD=1` is set at the same
time. Without the variable the step is the same as before.
- `doc/runner-image-contract.md` has a new section, "Choosing the Runner
for an image build". `doc/DEVELOPING.md` describes the new variable.
- The resolver is not an input of the Runner image, so the inputs digest
does not change.

## Verification

macOS (Node 26.12.0), in `server/`:

- `node node_modules/vitest/vitest.mjs run
scripts/prepare-runner-vendor.test.mjs
src/__tests__/runner-image-resolve.test.ts
src/__tests__/runner-image-publish.test.ts
src/__tests__/runner-image-identity.test.ts
src/__tests__/runner-image-workflow.test.ts
src/__tests__/server-package-build-script.test.ts`: 179 tests pass in 6
files. 27 of them are new resolver tests and 3 are new tests of the
vendor step, whose end-to-end test also has a new case.
- `tsc` on the new test file, with the settings of
`server/src/__tests__/tsconfig.json`: no error.

Linux (`node:24.22.3-bookworm`, GNU tar 1.34, the worktree mounted
read-only): the same files, 179 tests pass.

Real runs, on clean exports of the tree (`git archive`), with no
credential, on a `linux/arm64` Docker host:

- On `master` at `62d673cc3f`: `inputs` gave `lock=current` and
`inputs_digest=0e6662ad9c8c6451556c76bf326723645d4afaab2decc73cd5e0025f7dd8d1ca`.
`resolve` gave `mode=published`,
`image=ghcr.io/paperclipai/paperclip-runner@sha256:84f2c051aa4ed16c5cc138c08e3533fa417e92177f7d15af5f2aa3e0ba467162`
and
`slim_image=ghcr.io/paperclipai/paperclip-runner@sha256:f63ea43666ed708f1b0e5d2ade8bdc002885bf04252b9ed6cee21f7920e54f3c`.
These are the two indexes that `master` published for those inputs.
- The same tree with one line added to a Runner source file: another
digest, `mode=inline`, `reason=absent`.
- On `master` at `bc71c23e0a`, which changed the Runner's inputs, before
its image was published: digest
`88fe2929321b4706c77f11a098572ce990fec4449d505544c62ef3a84c28308a`,
`mode=inline`, `reason=absent`.
- Four deliberate faults in the resolver each make a test fail: no index
check, only `absent` is inline, the identity is built for a lockfile
that is behind, and the slim image is taken by tag.

Never run:

- A lookup with a registry credential. The package is public, so no
caller needs one now.
- Any workflow. No workflow calls the resolver in this pull request.
- `linux/amd64`.
- The full test suite, the full typecheck and `pnpm build`. `pnpm verify
--explain` reports exit 2, "broader verification required", so CI must
supply them.

## Risks

- Low for existing behaviour: nothing calls the resolver, and the vendor
step is the same unless the new variable is set.
- The resolver does not verify the attestation of the image. The
workflow that consumes the image must do that from the digest the
resolver reports.
- The resolver reports both build contexts, `runner-artifacts` and
`runner-artifacts-slim`. A build that reads only one stage gets one
context it does not use.
- A `broken` publication is `inline`, not a failure. The Runner image
workflow fails for that state.

> 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

- Claude Fable 5.2 (`claude-fable-5-2`), Anthropic, used through Claude
Code with tool use (shell, file edits, local Docker builds).

## 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

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Fable 5.2 <noreply@anthropic.com>
2026-11-12 22:03:15 -08:00
Nicky LeachandClaude Opus 5.5 ff8ec41e6a feat(runner): upgrade ACPX to 0.22.0 and move patched behavior into Paperclip (#20770)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - Paperclip Runner and the shared `adapter-utils` engine run ACP
agents through the embedded ACPX library
> - Paperclip pins ACPX 0.13.1 and keeps a large patch
(`patches/acpx@0.13.1.patch`) on it
> - ACPX 0.22.0 adds host hooks (`onAcpMessage`, `assertActive`,
`processLifecycle`, prompt-response `_meta`, `spawnAgent`-friendly
launch) that make much of that patch unnecessary
> - A large patch is expensive to port and hides behavior that belongs
to Paperclip
> - This pull request upgrades to ACPX 0.22.0 and moves most patched
behavior into Runner and `adapter-utils`
> - The benefit is a patch that is 484 lines instead of 1,334, with the
moved behavior covered by Paperclip tests

## Linked Issues or Issue Description

Refs #20675, #14430, #15075, #17942, #9980

**Proposed enhancement**
Upgrade the embedded ACPX library from 0.13.1 to 0.22.0. Keep only the
patch behavior that stock 0.22 cannot express.

**Motivation**
The 0.13.1 patch held behavior that Paperclip owns, for example usage
receipts, Copilot policy, model checks and the agent environment. Each
ACPX upgrade had to port all of it. ACPX 0.22 gives hooks that let
Paperclip do this work itself.

**Current behavior**
Runner and `adapter-utils` depend on 1,334 lines of patched ACPX 0.13.1
code.

**Expected behavior**
Runner and `adapter-utils` use ACPX 0.22.0. The patch keeps only:
extension methods, Cursor subagent routing, event `_meta` and the
message boundary chunk, the protocol guard hook, `spawnAgent`, and
`requestExtension`.

## What Changed

- ACPX 0.13.1 → 0.22.0 in Runner, `adapter-utils`, the server, the
Daytona runner images and the qualified profiles (`acpxVersion`).
`@agentclientprotocol/sdk` moves to 1.9.0.
- New `patches/acpx@0.22.0.patch` (484 lines). The 0.12.0 and 0.13.1
patches are removed.
- Delivery receipts: Runner confirms a permission reply when
`onAcpMessage` sees the exact outbound reply. Question and extension
replies count as delivered when Runner hands them back to ACPX. New
`interaction-delivery.ts`.
- Copilot policy: the existing policy function runs from `onAcpMessage`.
A violation stops the agent, cancels later permission requests and fails
the next prompt through `assertActive`.
- Model selection: stock 0.22 validation, plus a sidecar gate
(`createAcpxEffectiveModelGate`) that refuses a prompt if the agent
reports another model. Grok binds its model when the session opens.
Antigravity adds its concrete model to the model list ACPX reads.
- Typed session failures: Runner and `adapter-utils` read them from the
turn result `_meta` and fail the turn. They advertise the capability
themselves.
- Pi and Cursor usage receipts: a collector (`prompt-receipts.ts`) binds
them to the turn's own `session/prompt` reply.
- Goal capability: Runner reads it from the `initialize` reply. It keeps
it under `paperclip.goalCapability` in the saved agent capabilities
`_meta` and restores it when a saved session loads.
- Agent environment: Runner and `adapter-utils` set the agent
environment in `spawnAgent`. ACP terminals keep their current
environment. `adapter-utils` blocks inheritance of `XAI_API_KEY` and
refuses to start if the server has an `ACPX_AUTH_*` variable.
- `adapter-utils` spawn hooks (`spawnCwd`, stderr, `onAgentSpawn`): a
new `stockAcpRuntimeOptions` maps them to stock `spawnAgent` and
`processLifecycle`.
- Windows: ACPX 0.22 rejects raw command strings on Windows.
`adapter-utils` now passes an argv array there.
- Grok rules: a Runner guard (`grok-session-rules.ts`) writes
`session/new` `_meta.rules`.
- Turn-dispatch journal (#20675): ACPX 0.22 ignores errors from
`onAcpMessage`. The adapter keeps a failed `dispatched` write and throws
it from the turn's `assertActive`. ACPX calls `assertActive` before the
prompt write, so the write still stops. The real-ACPX test now checks
this route. `docs/durable-recovery.md` is updated.
- Plan events: the sidecar reads 0.22 `status` events tagged `plan`.
- CI: the backup-checkpoint lane now installs through the shared
`setup-ci-dependencies` action, like PR CI, so it resolves a stale
`pnpm-lock.yaml` in place. It runs on any root `package.json` change,
and pull requests never commit the lockfile, so its own frozen install
failed on dependency changes.
- Tests that checked removed patch internals are removed or rewritten:
the response-delivery, Cursor-usage and Pi-receipt package contracts,
and the goal-persistence packaging test. New unit tests cover the moved
behavior.

## Verification

Local runs on macOS (darwin-arm64), Node 26, on the rebased head:

- `pnpm --filter @paperclipai/paperclip-runner exec vitest run
src/drivers/acpx src/cli`: 2,000 passed.
- `tsc -p packages/paperclip-runner/tsconfig.json` and
`tsconfig.test.json`: no errors, except the existing
`@paperclipai/paperclip-eval-kernel` build gap in `workflow-harness.ts`.
- Runner package contracts (`node --test
test/acpx-*-package-contract.test.mjs
test/acpx-sidecar-contract.test.mjs`), `scripts/build-inputs.test.mjs`,
`scripts/contracts/contract-fingerprints.test.mjs`,
`scripts/acpx-patch-packaging.test.mjs`: all pass. `node
scripts/generate-acpx-sidecar-contract.mjs --check` passes.
- `cargo test -p paperclip-runner-core --test acpx_park_preparation
--test acpx_park_readiness --test acpx_warm_attachment --test
native_provider_backend`: 86 passed. `cargo fmt --check` passes.
- `adapter-utils`: `tsc` clean. `vitest run src/acpx-engine
src/mcp-isolation.integration.test.ts`: 457 passed (one Gemini flag test
failed once under load and passes when run alone).
- Adapters: `claude-local` and `codex-local` `acp.quota.test.ts` (23 and
39 passed), `grok-local` (149 passed).
- Server: `scripts/prepare-runner-vendor.test.mjs`,
`native-session-executor.test.ts`, `quarantined-task-retry.test.ts`
(1,416 passed).
- `cargo test -p paperclip-runner-core --lib -- previous_acpx_release
acpx_upgrade unrestorable_saved_sessions`: 3 passed. `node --test
test/acpx-session-carry-forward-package-contract.test.mjs`: passes.
- `patch -p1 --forward` on a clean `acpx-0.22.0.tgz` gives output
identical to the port.
- Under full-suite load, some Runner live and recovery tests time out.
Each one passes when run alone. Full regression is left to CI.

Not run: live provider runs (they need provider credentials), the opt-in
Kimi live peer test, the Gemini filesystem smoke test, and Windows.

## Risks

- Previous-release story: sessions saved by the previous release (ACPX
0.13.1) carry forward. ACPX 0.22 loads the session records 0.13.1 wrote.
Runner moves a saved descriptor that names 0.13.1
(`CARRIED_ACPX_RELEASES`) to 0.22.0, and the sidecar keeps 0.13.1 as the
ACPX value in the session identity, so the saved session key and
identity record still match. The existing launch-profile rebind still
decides which sessions resume: a settled session, or the same run's
interrupted turn through lost-turn recovery. A session saved by any
other ACPX release is still refused as `provider_state_incompatible`.
`rules/acpx-profile-identity.mdc` and `docs/durable-recovery.md` now
describe this.
- Proof: `acpx-session-carry-forward-package-contract.test.mjs` resumes
a session record written by the real ACPX 0.13.1 with its Paperclip
patch. 0.22 loads the same provider session (no `session/new`), runs a
turn and keeps the earlier turn. Rust tests cover a settled 0.13.1
session that the next run attaches to, and a turn interrupted by the
upgrade. A recovery-identity test checks that the session key and
identity record match across the versions.
- `pnpm-lock.yaml` is not committed. CI resolves it for each run, and
the lockfile refresh workflow updates it after merge.
- Behavior change on the server: `adapter-utils` refuses to create an
ACPX runtime if the server environment has an `ACPX_AUTH_*` variable.
Before, ACPX ignored such variables. Stock 0.22 would use them to
authenticate agents with server credentials.
- Delivery confirmation is now Runner-owned. A permission reply is
confirmed by the exact outbound message. Question and extension replies
are confirmed when Runner hands them to ACPX, not when the pipe accepts
them.
- Users can select only models that the agent advertises. Stock 0.22
rejects other model IDs (Claude is exempt).
- ACP terminals keep the sidecar or server environment, as before. This
existing exposure is not changed by this pull request.
- The `adapter-utils` Gemini version probe now runs with the server
environment plus the run variables, not the run variables alone.
- The pack contract digest and profile `acpxVersion` change, so Daytona
needs a new runner image.
- Grok launch flags (`--model`, `--reasoning-effort`) were checked on
1.0.13 but session open needs a login, so Grok keeps the existing
`_meta` path.

> 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

Claude Opus 5.5 (`claude-opus-5-5`), Anthropic, in Claude Code with tool
use (shell, file editing, subagents). Extended reasoning was used.

## 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

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-11 07:28:00 -08:00
7df517d18f perf(build): reuse a current Runner build in the server's prepare:runner-vendor step (#15926)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - Its CI runs a `Build` job on every pull request. That job runs `pnpm
build` and then verifies the public npm install. Since the server test
shards got faster, `Build` is the slowest check in most runs (419-497s
on the six fully green `master` runs created 2026-10-11 19:46Z to
20:17Z).
> - Inside `pnpm build`, the server package is the last package to
build. Its first step, `prepare:runner-vendor`, runs the full build of
`@paperclipai/paperclip-runner` so the server can vendor
`packages/paperclip-runner/dist`. In `pnpm -r build` that package was
built seconds earlier by the same command. The second build took 12-16s
on those six runs and sits on the critical path of the job.
> - The step cannot just be removed. A standalone `pnpm --filter
@paperclipai/server build` (the Dockerfile's path) must produce a
current vendored Runner on its own.
> - This pull request keeps that guarantee and skips only the build that
is provably redundant. The Runner's `build` records a digest of its
inputs, captured before its first step and confirmed after its last,
together with a manifest of every file it left under `dist`. The
server's `prepare:runner-vendor` reuses the build when the digest
matches, the stamp is for this platform, and every recorded output is
still there at its recorded size. Any other state runs the same full
Runner build as before.
> - The benefit is 12-16s off the `Build` job and off every other `pnpm
build` (the Canary Dry Run's `release.sh` build, local builds), with no
change to what a standalone server build produces. The PR's own first
run measured it: the server step reused the build in 56ms where `master`
runs rebuilt for 11-15s.

## Linked Issues or Issue Description

No GitHub issue exists for this. The description below follows the
enhancement template.

**What existing behavior does this improve?**

The server build's `prepare:runner-vendor` step, which builds
`@paperclipai/paperclip-runner` so the server can vendor its `dist`.

**Subsystem affected**

Build and CI (`server/package.json`,
`packages/paperclip-runner/package.json`, the `Build` job).

**Current behavior**

`prepare:runner-vendor` runs the Runner's full `build` every time.
Inside `pnpm -r build`, the Runner was already built by the same command
20 to 25 seconds earlier, with every check passing. On six fully green
`master` runs (2026-10-11, 19:46Z to 20:17Z) the repeated build took
11.8s to 15.9s, measured from the `server build$ pnpm run
prepare:runner-vendor` log line to the Runner's `Finished release
profile` line. The server package builds last, so that time is on the
critical path of `Build`, which is the slowest check in most pull
request runs.

**Proposed behavior**

- The Runner's `build` starts with `node scripts/record-build-inputs.mjs
--begin` and ends with `node scripts/record-build-inputs.mjs --finish`.
`--begin` removes the previous stamp and records a digest of every build
input before the first step reads anything. `--finish`, after every step
and check has passed, recomputes the digest and writes
`packages/paperclip-runner/node_modules/.cache/paperclip-runner-build-inputs.json`
only if the inputs are still the ones the build began with, together
with the path and size of every file under `dist`. A stamp therefore
exists only for a build that completed every step on inputs that did not
move while it ran.
- The server's `prepare:runner-vendor` becomes `node
scripts/prepare-runner-vendor.mjs`. It reuses the build if the stamp is
for this platform and architecture, every recorded output is still there
at its recorded size, and the recomputed digest matches. In every other
case it removes the stamp and runs `pnpm --filter
@paperclipai/paperclip-runner build`, the exact command the step ran
before.
- `PAPERCLIP_RUNNER_VENDOR_REBUILD=1` forces the rebuild.

**Reason and benefit**

12-16s off the `Build` job and off every other `pnpm build` (the Canary
Dry Run's `release.sh` build, local builds), with no change to what a
standalone `pnpm --filter @paperclipai/server build` produces.

**Breaking changes**

None. The rebuild path runs the same command as before. A fresh
checkout, a changed input, a missing or truncated output, an edit that
landed during the build, or a stamp from another platform always
rebuilds.

## What Changed

- `packages/paperclip-runner/scripts/build-inputs.mjs` (new): defines
the input set, the digest, the output manifest and the reuse decision.
Inputs are every file under `packages/paperclip-runner` except `dist`,
`dist-browser`, `runner/target`, `node_modules` and dot-directories;
every file under `packages/paperclip-eval-kernel` except `dist` and
`node_modules` (the Runner's `build:typescript` builds it first); every
file under the root `patches` directory (`check:acpx-profiles` hashes
`acpx@0.13.1.patch`, and `build-verified-provider-entrypoints.mjs`
copies the `pi-acp` and `brace-expansion` patches into
`dist/cli/pi-provision-inputs`); and the root `package.json`,
`pnpm-lock.yaml`, `pnpm-workspace.yaml` and `tsconfig.base.json`. The
toolchain pins are inside the package (`rust-toolchain.toml`,
`runner/Cargo.lock`) and the TypeScript version is in the lockfile, so
both are covered. A symlink is hashed as its target path and is not
followed. Outputs are every file under `dist`, the tree the server
copies wholesale, recorded with their sizes. Exports `beginRunnerBuild`,
`recordRunnerBuild` and `assessRunnerBuild` (returns `current` plus a
stable reason code: `no_stamp`, `schema_mismatch`, `platform_mismatch`,
`missing_output`, `outputs_changed`, `inputs_changed`,
`inputs_unchanged`).
- `packages/paperclip-runner/scripts/record-build-inputs.mjs` (new): the
two phases, `--begin` and `--finish`, added as the first and last step
of the Runner's `build` in `packages/paperclip-runner/package.json`.
Each build chain has its own pending record under
`node_modules/.cache/paperclip-runner-build-inputs.pending/`, named
after the shell pnpm runs the chain in (the parent of both phases;
`--finish` also accepts a record named after its own pid, for a shell
that execs the last command of the chain), so two Runner builds running
at once in one checkout do not consume each other's record. `--finish`
without a pending record for its chain is an error, so the chain cannot
silently lose its first step; `--finish` after an edit that landed
mid-build prints `not recorded: inputs changed while the build ran` and
leaves no stamp. Records of builds that never finished are pruned by the
next `--begin` after a day.
- `server/scripts/prepare-runner-vendor.mjs` (new): the reuse-or-rebuild
step. It resolves the pnpm that is running it through `npm_execpath`
with the same rule as the CLI's `resolvePnpmInstallInvocation` (a
JavaScript entry runs through this Node, a native executable runs
directly, otherwise `pnpm` on PATH). It prints one
`paperclip.runner-vendor.timing.v1` line with `action`, `reason`,
`check_ms` and `build_ms`, so the saving is measured on this PR's own
`Build` job.
- `server/package.json`: `prepare:runner-vendor` now runs that script.
The `build` and `typecheck` scripts are unchanged.
- `server/src/__tests__/server-package-build-script.test.ts`: pins the
new indirection and the command behind it (imported from the script, so
the Dockerfile's standalone path stays pinned), and a new test asserts
the Runner's `build` starts with `--begin`, ends with `--finish`, and
has exactly those two record steps.
- Tests: `packages/paperclip-runner/scripts/build-inputs.test.mjs`
(node:test, added to `test:typescript:prep`) and
`server/scripts/prepare-runner-vendor.test.mjs` (vitest, in the server
lane).
- `doc/DEVELOPING.md`: a "Reused Runner build in the server build"
section under Test Commands.

## Review findings on the first head, all fixed in
[2d57542](https://github.com/paperclipai/paperclip/pull/15926/commits/2d575425977f87b4612222d0a1961bc9cc5ba569)

Greptile scored the first head 2/5 with three P1 findings. All three
were valid.

| Finding | Fix |
| --- | --- |
| **Patch edits leave stale output.** The root patch files the build
reads were not in the digest. | Every file under the root `patches`
directory is an input. Tests assert the patches are in the input list
and that editing the pi-acp patch, the acpx patch, or adding a patch
changes the digest. |
| **Missing helpers pass the build.** Only four outputs were checked, so
a `dist` that lost `dist/cli/acpx-runtime-sidecar.cjs` still counted as
current. | `--finish` records the path and size of every file under
`dist`; reuse requires all of them intact and reports `outputs_changed`
naming the first missing or resized file. Schema bumped to
`paperclip.runner-build-inputs.v2` so an older stamp is rejected. Tests
cover a deleted helper, a truncated helper, and the server step
rebuilding on `outputs_changed`. |
| **New edits count as compiled.** The digest was hashed after the
build, so a source edited while `build:binary` ran was recorded as
compiled. | The build is bracketed: `--begin` captures the digest before
the first step, `--finish` compares after the last and writes nothing if
it moved. Tests cover the mid-build edit, the undisturbed build, and the
CLI in both orders. |

## Review finding on the second head, fixed in the third commit

Greptile scored the second head 4/5 with one P1, valid: two Runner
builds running at once in one checkout (server `build` beside
`typecheck`) shared one pending record, so the second `--begin`
overwrote it, the first `--finish` consumed it, and the second
`--finish` threw `no pending build inputs` after every step of its build
had passed. Reproduced on the second head's module: two
`beginRunnerBuild`, then the first `recordRunnerBuild` records and the
second throws.

| Finding | Fix |
| --- | --- |
| **Overlapping builds fail each other.** One pending file for every
build chain. | One record per chain, named after the shell pnpm runs the
chain in: `--begin` sees it as its parent, and so does `--finish`,
unless the shell execs the last command of the chain (bash can), in
which case `--finish` is that shell and finds the record under its own
pid. Pending records of builds that never finished are pruned after a
day. An edit between two overlapping starts leaves no stamp whichever
build finishes last: the build that began before the edit reports
`inputs_changed_during_build` and removes the stamp the other wrote, so
a stamp always means every build that wrote `dist` read the inputs it
names. Tests: both finishing orders with and without an edit, the
pruning, a `--finish` that cannot use another chain's record, and two
real `sh -c` chains (plus one under `bash`) held open at the same time,
both recording. On the real repo, two `pnpm --filter
@paperclipai/paperclip-runner build` chains started 8s apart with a
cargo stand-in both recorded (`files=1223 outputs=1434`) and left the
pending directory empty. |

## Measured on this PR's runs

The `Build` job ran `pnpm build` from the branch, so both builds of the
Runner in that job used the new chain.

First head (c23faab, run 38175281058), against the four fully green
`master` runs created 20:41Z to 21:37Z the same evening:

| | PR run 38175281058 | `master`, four runs |
| --- | --- | --- |
| Runner build inside `pnpm -r build` | 55.3s | 47.3-62.0s |
| Server step `prepare:runner-vendor` |
`action:reuse,reason:inputs_unchanged,check_ms:56,build_ms:0` | full
Runner rebuild, 11.0-14.8s |
| Server package build | 29.7s | 32.3-43.5s |
| `pnpm build` step | 114s | 102-135s |
| `Build` job | 411s | 409-474s |

Second head (2d57542, run 38178686999), against the two fully green
`master` runs created in the same half hour (38178951607 and
38180574768):

| | PR run 38178686999 | `master`, two runs |
| --- | --- | --- |
| Runner build inside `pnpm -r build` | 53.9s (`begin:` 22:03:04,
`recorded:` 22:03:57, digest 460e02616081 over 1231 files and 1434
outputs) | 62.2s, 64.6s |
| Server step `prepare:runner-vendor` |
`action:reuse,reason:inputs_unchanged,check_ms:109,build_ms:0` | full
Runner rebuild, 14.5s, 15.2s |
| Server package build | 27.4s | 43.5s, 44.8s |
| `pnpm build` step | 106s | 91s, 95s to the Runner's end plus the rest;
127s, 135s in all |
| `Build` job | 405s | 461s, 479s |

The rebuild is gone from the server build on both heads, and the check
costs 56-109ms where `master` rebuilds for 11-15s; the rest of the
spread between runs is runner speed on the `runs-on` fleet. The third
commit changes nothing on the reuse path: it only changes where the
pending record lives between `--begin` and `--finish`.

## Verification

- `node --test packages/paperclip-runner/scripts/build-inputs.test.mjs`:
11 tests, three consecutive runs clean. Two builds at once in one
checkout each keep their own pending record and both record, in either
finishing order; with an edit between their starts the build that began
before it leaves no stamp whichever finishes last. A build's start
prunes records older than a day and no others. A `--finish` cannot use
another chain's record. Two chains run by separate `sh -c` shells at the
same time (and one under `bash`) both record and leave the pending
directory empty. The input set is exactly the sources, the eval kernel,
the root patches and the root config, and skips outputs and caches. The
output manifest lists every file under `dist` with its size and nothing
outside it. The digest is stable across reads, does not change when
`dist`, `dist-browser`, `node_modules`, `.vite` or `runner/target`
change, and does change for a Runner source, a Rust source, an
eval-kernel source, the pi-acp patch, the acpx patch, a new patch, the
lockfile, the base tsconfig, a new file and a rename. A symlink changes
the digest but its target's files are not read. A build is current only
with a matching v2 stamp, the same platform and architecture, the four
named outputs present, and every recorded output intact (a deleted or
truncated helper is `outputs_changed`; a file the build did not produce
may appear). An input edited between `--begin` and `--finish` leaves no
stamp and removes the pending record; the next build records normally.
`--finish` without `--begin`, or after a build that left a required
output missing, throws and leaves no stamp. The CLI brackets the build
for the package it is pointed at, reports a mid-build edit, and rejects
`--finish` first, both flags, neither flag, and unknown arguments.
- `vitest run scripts/prepare-runner-vendor.test.mjs
src/__tests__/server-package-build-script.test.ts` in `server/`: 21
tests. The decision reuses a current build, rebuilds a stale one, and
rebuilds a current one only when `PAPERCLIP_RUNNER_VENDOR_REBUILD=1`.
The pnpm resolver covers a JavaScript entry, a native executable, a
Windows executable, no `npm_execpath`, and npm's `npm_execpath`. With an
injected build command: a current stamp runs nothing and prints
`action:reuse`; changed inputs run the exact command from `server/` with
the stale stamp already removed; a lost
`dist/cli/acpx-runtime-sidecar.cjs` rebuilds with
`reason:outputs_changed`; a failing build returns its status and leaves
no stamp. End to end, the CLI with a stand-in `pnpm` on PATH that
brackets its build the way the real one does rebuilds on the first call,
reuses on the second, rebuilds after a source change, rebuilds after a
lost output, and reuses again. The guard test asserts
`prepare:runner-vendor` is the new script, the rebuild command is still
`pnpm --filter @paperclipai/paperclip-runner build`, the server build
still starts with the step, and the Runner's `build` starts with
`--begin` and ends with `--finish` with exactly two record steps; it
fails against the previous build script on its first assertion.
- Real repo, this checkout, with `cargo` and `pnpm` stand-ins on PATH
(no Rust toolchain here; the stand-in writes the one file
`stage-runner-binary.mjs` copies): `pnpm --filter
@paperclipai/paperclip-runner build` ran the complete chain through the
real nested pnpm (every TypeScript step and check, the stand-in cargo,
the binary staging, the two binary-reading checks) and recorded
`files=1223 bytes=18735059 outputs=1434`; the stamp is 128 KB. `pnpm run
prepare:runner-vendor` in `server/` then printed
`action:reuse,reason:inputs_unchanged,check_ms:222,build_ms:0`.
Appending a newline to `patches/pi-acp@0.0.33.patch` made the assessment
`inputs_changed`; truncating or deleting
`dist/cli/acpx-runtime-sidecar.cjs` or
`dist/cli/opencode-app-server-proxy.cjs` made it `outputs_changed` in
1-5ms; restoring each made it current again. `--begin`, an edit to
`src/index.ts`, then `--finish` printed `not recorded: inputs changed
while the build ran` and left no stamp.
- The whole check on the real set (1223 inputs, 18 MB; 1434 outputs)
takes 165-227ms on this host and took 56ms on the CI runner; the output
manifest part is 1-5ms.
- Not run here: the real `cargo build`, and `pnpm build` end to end.
This PR's `Build` job runs `pnpm build` from the branch, so its log
carries the Runner's `begin:` and `recorded:` lines and the server
build's `paperclip.runner-vendor.timing.v1` line with the measured
result.

## Risks

- A build input outside the digest would let a stale Runner be vendored.
The digest covers the whole Runner package except its outputs, the eval
kernel the Runner builds first, the root patches, and the root files the
Runner's tsconfig and dependency graph read. The lockfile covers
dependency versions, and the pins inside the package cover the Rust
toolchain. An input missed here is a bug to fix by adding it to the set;
the fallback direction is always a rebuild, never a skip.
- The whole `patches` directory is an input, although the Runner build
reads three of its fourteen files. An edit to an unrelated patch
therefore costs one extra Runner rebuild, which is the safe direction
and avoids a hand-kept list of file names that could drift from the
scripts that read them.
- The stamp lives in `node_modules/.cache`, which pnpm does not manage
and `.gitignore` and `.dockerignore` already exclude. It is about 128 KB
with the output manifest. The Docker build has no stamp and no prior
Runner build, so it rebuilds exactly as before.
- A stamp is written only after the Runner's `build` passed every step
on inputs that did not change while it ran. The Runner's first step and
the prepare step both remove any existing stamp before a rebuild starts,
so a build that stops partway, or a build command that fails before its
first step, leaves no stamp.
- `node scripts/record-build-inputs.mjs --finish` on its own is an
error; the script is only meant to run as the two ends of the `build`
chain, and the guard test pins that.
- The pending record is keyed by the pid of the shell that runs the
chain. pnpm runs every script through a shell (`sh -c` on POSIX, `cmd`
on Windows), and both phases are children of that shell; a shell that
execs the last command of the chain is covered by matching the finishing
process's own pid too. A chain run some other way, with the two phases
under different parents, fails loudly at `--finish` rather than
recording. Two builds running at once still write the same `dist`, which
no stamp can make safe; what the record guarantees is that a stamp is
written only by a build whose own inputs did not move.
- The platform and architecture are part of the comparison, so a stamp
from another machine is never trusted; an older v1 stamp is rejected by
schema.
- The Typecheck job builds the Runner once, through this step, with no
prior `pnpm -r build`, so it gains nothing and loses nothing: there is
no stamp and it rebuilds as before.
- The step now spawns pnpm through `npm_execpath` instead of relying on
`pnpm` being on PATH, because under corepack the script shell has no
`pnpm`. Outside a pnpm script it still falls back to `pnpm` on PATH.

## Model Used

- Claude (Anthropic), model id `claude-fable-5-1` (Claude Fable 5.1),
extended thinking, with tool use (shell, file editing, GitHub CLI) in a
Paperclip agent 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
- [ ] 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

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: devinfoley <139239+devinfoley@users.noreply.github.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
2026-10-12 11:25:24 -07:00