SRE-984: Move Turborepo from generated package.json to Cargo workspaces

[TimDiekmann]

🌟 What is the purpose of this PR?

Turborepo reads the Cargo workspace natively behind futureFlags.experimentalCargoWorkspaces. Every crate becomes a package under its Cargo.toml name, its dependencies follow Cargo.toml, and [workspace.metadata].name names the package that stands for the whole workspace. The generated package.json mirrors and sync-turborepo, which existed only to describe the crates to turbo, go away.

turbo prune also learns the Cargo workspace: it writes the pruned Cargo.toml, Cargo.lock and Cargo configuration itself and follows dev-dependencies. That retires the dummy crate manifests the Dockerfiles generated and most of the prune action's own scope computation.

🔗 Related links

• SRE-984
• Rust RFC discussion, including the report of the dependencyOutputs issue below

🚫 Blocked by

• SRE-1061: Move the crates' package.json scripts into turbo.json commands #9755

🔍 What does this change?

• Enables experimentalCargoWorkspaces; the root Cargo.toml names the workspace hash
• Deletes the 52 crate package.json files and sync-turborepo: the repo-chores module and CLI subcommand, the mise task, the root script, the lint step, and the [package.metadata.sync.turborepo] sections
• Refers to the crates by their Cargo names everywhere: the six TypeScript consumers' task edges, the workflows, the Dockerfile scopes, the crates' own turbo.json commands, the changeset config, and the docs
• The four TypeScript packages and the hash-graph-integration crate that depended on @apps/hash-graph depend on hash-graph#start:test:healthcheck instead, which is what they need from it
• The task documents cover the tasks a package implements — a package.json script or a command in a turbo.json — so turbo's built-in Cargo tasks (build, check, lint, test, format) stay out of them; the generator no longer needs packages { tasks }
• The prune action calls turbo prune directly; prune.py and its closure over both graphs are gone. It merges the root dot-files into the pruned tree and copies the two Brunch directories no package owns; the Petrinaut docs arrive through a manifest edge on @apps/brunch-agent#test:unit
• The Dockerfiles prune without stubbing workspace members. Turbo runs cargo metadata whenever it loads the graph, so the prune stages carry the Rust toolchain, as does the petrinaut-opt bindings stage
• The five dependencyOutputs blocks are removed in their own commit while turbo's lazy native discovery rejects dependencyOutputs.from selectors that point at a crate outside the filter (reported in the discussion above); the commit is reverted once the fix ships
• The nextest override for the retried harpc-net tests matches by test name: nextest validates package matchers against the workspace, and a pruned tree without the crate rejects them
• The deploy catalog and the workflows name the graph image's package hash-graph
• tests/graph/http is named @tests/hash-graph-http

Pre-Merge Checklist 🚀

🚢 Has this modified a publishable library?

This PR:

• does not modify any publishable blocks or libraries, or modifications do not need publishing

📜 Does this require a change to the docs?

The changes in this PR:

• require changes to docs which are made as part of this PR

🕸️ Does this require a change to the Turbo Graph?

The changes in this PR:

• affected the execution graph, and the turbo.json's have been updated to reflect this

⚠️ Known issues

• Turbo registers build, check, lint, test and format for every crate (run and dev for binaries) with plain Cargo commands. Nothing runs them unfiltered in CI, and a crate's own command for a name wins over the built-in one, so they stay; opting out would cost five entries in each of the 52 crate configurations.
• The prune action still copies libs/@hashintel/brunch-agent/docs and libs/@hashintel/brunch-agent/evaluations: @apps/brunch-agent reads them by path, but they sit next to the Brunch packages rather than inside one, so no edge can carry them into a pruned tree. Giving each directory a package.json the app depends on would turn them into ordinary edges; that is a Brunch change.
• The Bench workflow prunes the base branch with the package names of the head. While the base is SRE-1061: Move the crates' package.json scripts into turbo.json commands #9755, hash-graph-benches does not exist there and the base benches fail; that resolves itself once the base carries this change.
• The three Python packages keep their package.json mirrors; turbo's experimentalPythonWorkspaces needs a uv workspace, which is a separate step.

🐾 Next steps

• Revert the dependencyOutputs removal once turbo loads crate scopes before validating dependencyOutputs.from
• Move the Python packages onto experimentalPythonWorkspaces once they share a uv workspace

🛡 What tests cover this?

• The prune action was replayed locally for hash-graph-postgres-store, @apps/hash-api, @apps/hash-frontend and @apps/brunch-agent: cargo metadata --locked loads every pruned workspace, the background tasks match today's. Every crate task resolves the same command as before; the task documents of the JavaScript packages change only in the crate names they reference.
• CI on this branch: Lint, Test and Deploy (all eight images) pass on pruned trees.

❓ How to test this?

1. Checkout the branch, mise install --locked npm:turbo
2. turbo ls lists the crates under their Cargo names and hash for the workspace
3. turbo prune hash-graph --docker --out-dir /tmp/prune && cd /tmp/prune/full && cargo metadata --locked succeeds without a stub in sight
4. turbo run lint:clippy --filter error-stack runs just clippy from the crate's turbo.json

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

4 Skipped Deployments

Project	Deployment	Actions	Updated
hash	Ignored	Preview	Sep 16, 2026 2:50pm UTC
hashdotdesign-tokens	Ignored	Preview	Sep 16, 2026 2:50pm UTC
petrinaut	Skipped		Sep 16, 2026 2:50pm UTC
petrinaut-docs	Skipped		Sep 16, 2026 2:50pm UTC

[github-advanced-security]

You are seeing this message because GitHub Code Scanning has recently been set up for this repository, or this pull request contains the workflow file for the Code Scanning tool.

What Enabling Code Scanning Means:

• The 'Security' tab will display more code scanning analysis results (e.g., for the default branch).
• Depending on your configuration and choice of analysis tool, future pull requests will be annotated with code scanning analysis results.
• You will be able to see the analysis results for the pull request's branch on this overview once the scans have completed and the checks have passed.

For more information about GitHub Code Scanning, check out the documentation.

[codspeed]

Merging this PR will not alter performance

⚠️ 6 benchmarks measured no execution time

Nothing ran under measurement, usually because the compiler removed the code under test. These results are not comparable, so they count as unchanged.

Preventing compiler optimizations

✅ 98 untouched benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚠️	as_constant	< 1 ns	< 1 ns	N/A
⚠️	constant_equal	< 1 ns	< 1 ns	N/A
⚠️	constant_not_equal	< 1 ns	< 1 ns	N/A
⚠️	access	< 1 ns	< 1 ns	N/A
⚠️	runtime_equal	< 1 ns	< 1 ns	N/A
⚠️	runtime_not_equal	< 1 ns	< 1 ns	N/A

---

_{Comparing t/sre-984-move-turborepo-from-generated-packagejson-to-cargo (77a8e70) with t/sre-1061-move-the-crates-packagejson-scripts-into-turbojson-commands (2cd552e)}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 66.28%. Comparing base (2cd552e) to head (8fdc3b8).

Additional details and impacted files

@@                                           Coverage Diff                                           @@
##           t/sre-1061-move-the-crates-packagejson-scripts-into-turbojson-commands    #9758   +/-   ##
=======================================================================================================
  Coverage                                                                   66.28%   66.28%           
=======================================================================================================
  Files                                                                        1775     1775           
  Lines                                                                      191861   191861           
  Branches                                                                     7853     7853           
=======================================================================================================
+ Hits                                                                       127171   127174    +3     
+ Misses                                                                      63209    63206    -3     
  Partials                                                                     1481     1481           

Flag	Coverage Δ
antsi	2.36% <ø> (?)
apps.hash-ai-worker-ts	1.99% <ø> (ø)
apps.hash-api	15.35% <ø> (ø)
blockprotocol.type-system	38.15% <ø> (ø)
error-stack	90.81% <ø> (?)
harpc-codec	84.70% <ø> (?)
harpc-net	96.24% <ø> (?)
harpc-tower	67.03% <ø> (?)
harpc-types	0.00% <ø> (?)
harpc-wire-protocol	92.23% <ø> (?)
hash-codec	72.76% <ø> (?)
hash-config	81.14% <ø> (?)
hash-graph	12.54% <ø> (?)
hash-graph-api	19.71% <ø> (?)
hash-graph-atlas	80.36% <ø> (?)
hash-graph-authentication	96.02% <ø> (?)
hash-graph-authorization	63.14% <ø> (?)
hash-graph-embeddings	91.88% <ø> (?)
hash-graph-postgres-store	32.15% <ø> (?)
hash-graph-store	48.41% <ø> (?)
hash-graph-temporal-versioning	50.18% <ø> (?)
hash-graph-types	0.00% <ø> (?)
hash-graph-validation	84.71% <ø> (?)
hash-middleware	90.92% <ø> (?)
hashql-ast	89.63% <ø> (?)
hashql-compiletest	28.39% <ø> (?)
hashql-core	78.92% <ø> (?)
hashql-diagnostics	72.51% <ø> (?)
hashql-eval	79.82% <ø> (?)
hashql-hir	89.09% <ø> (?)
hashql-mir	87.92% <ø> (?)
hashql-syntax-jexpr	94.04% <ø> (?)
local.claude-hooks	0.00% <ø> (ø)
local.harpc-client	51.49% <ø> (ø)
local.hash-backend-utils	3.27% <ø> (ø)
local.hash-graph-sdk	10.02% <ø> (ø)
local.hash-isomorphic-utils	12.22% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.