[rust] worker interface

[mikea]

upstreaming internal work to start building on it.

[ask-bonk]

I'm Bonk, and I've done a quick review of your PR.

This introduces a Rust worker::Interface trait mirroring workerd::WorkerInterface, with FFI bridging and two simple implementations (error worker, kill-switch worker).

Findings (highest severity first):

1. [MEDIUM] SCRIPT_KILLED_DETAIL_ID in bridge.h uses const uint64_t instead of constexpr kj::Exception::DetailTypeId, inconsistent with the rest of the codebase.
2. [LOW] error.rs:86 uses std::result::Result<AlarmResult, KjError> while all other methods in the same file use crate::Result<T> — inconsistent.
3. [LOW] Commented-out internal references leaked from internal repo (edgeworker namespace in test, commented-out #include and BUILD dep).
4. [LOW] Typo in lib.rs doc comment: "rusts's" should be "Rust's".

---

These comments were generated by an AI code review assistant and may contain inaccuracies.

[ask-bonk]

Review posted successfully on PR #6621 with 5 inline comments covering:

1. [MEDIUM] SCRIPT_KILLED_DETAIL_ID should use constexpr kj::Exception::DetailTypeId to match the codebase convention (with suggestion).
2. [LOW] Inconsistent std::result::Result vs crate::Result in error.rs (with suggestion).
3. [LOW] Commented-out internal includes/namespace in test file.
4. [LOW] Commented-out internal BUILD dependency.
5. [LOW] Typo "rusts's" in doc comment (with suggestion).

github run

[ask-bonk]

I'm Bonk, and I've done a quick review of your PR.

This PR introduces a Rust worker::Interface trait mirroring workerd::WorkerInterface, with FFI bridging via CXX, and two simple implementations (error worker, kill-switch worker).

Findings (highest severity first):

1. [HIGH] CPU_LIMIT_DETAIL_ID value mismatch between Rust and C++ — the Rust constant has a missing hex digit, causing a silently wrong value.
2. [LOW] compile_flags.txt references errors.rs@cxx but the source file is error.rs (no trailing 's').
3. [LOW] All new files (.rs, .h, .c++) are missing copyright/license headers, which the codebase requires.

Note: the SCRIPT_KILLED_DETAIL_ID const/constexpr issue from the previous review is still valid and unaddressed.

---

These comments were generated by an AI code review assistant and may contain inaccuracies.

[ask-bonk]

Review posted successfully on PR #6621 with 3 inline comments covering:

1. [HIGH] CPU_LIMIT_DETAIL_ID value mismatch — the Rust constant 0xfdc_b787_ba24_0576 is 15 hex digits, missing a 4 compared to the C++ value 0xfdcb787ba4240576ull (16 hex digits). These are entirely different values. (with suggestion)
2. [LOW] compile_flags.txt typo: errors.rs@cxx should be error.rs@cxx (with suggestion)
3. [LOW] All 8 new files are missing copyright/license headers required by codebase convention.

github run

[ask-bonk]

UnknownError: ProviderInitError

github run

[ask-bonk]

@mikea Bonk workflow failed. Check the logs for details.

View workflow run · To retry, trigger Bonk again.

[ask-bonk]

UnknownError: ProviderInitError

github run

[ask-bonk]

LGTM

github run

[codspeed-hq]

Merging this PR will not alter performance

✅ 72 untouched benchmarks
⏩ 129 skipped benchmarks¹

---

_{Comparing maizatskyi/2026-04-20-rust-worker (861eb04) with main (94328e1)}

Footnotes

1. 129 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[jasnell]

Couple of nits, otherwise LGTM