docs(crate): ENG-4683 README, axum quickstart, and architecture diagram by jimbofreedman · Pull Request #20 · spotship/rustyroute

jimbofreedman · 2026-07-27T10:19:55Z

Summary

Turns README.md from a repository-skeleton note into an OSS-facing
document, and — more importantly — makes it executable so it cannot
rot against the API. Closes ClickUp ENG-4683.

The README now carries the eleven sections the ticket asked for: badges,
the pitch, a pre-1.0 status callout, a library quickstart, a ~50-line
axum HTTP server, the build-pipeline diagram, the feature/size table,
the 13 edge groups, performance, licence/attribution, and contributing.
The existing CLI section is preserved (see Deliberate deviations).

What makes it more than a docs change: nothing in the README was
compiled by anything before this PR. A #[cfg(doctest)] anchor in
src/lib.rs now feeds the README's fences to cargo test --doc, so the
library quickstart is compiled and executed on every CI run, and
examples/axum_server.rs is the single source of truth for the HTTP
block — the README embeds the same bytes and a test asserts they are
byte-identical.

Every factual claim is machine-checked or measured. The 13-group table
is asserted element-for-element against the generated
rustyroute::EDGE_GROUPS (names and their upstream pass tags, in
order); the menaiStrait bbox is asserted against build/groups.rs;
the section inventory is asserted as an ordered vector. The archive
sizes, node/edge counts and latency figures were measured on this
worktree, not estimated, and budgets are kept separate from observations.

Deliberate deviations from the ticket

The CLI section is kept. The ticket says the README "replaces the
current placeholder" and lists eleven sections, none of them CLI. But
that section documents shipped ENG-4682 behaviour and I re-verified it
is still accurate (flags, defaults, exit codes 0–4). Deleting correct
user-facing docs to satisfy a literal reading would be a regression.
Easy to drop if you disagree.
tokio gets a net feature the ticket didn't list. The ticket
specifies features = ["macros", "rt-multi-thread"], which cannot
compile the example: axum::serve takes a tokio::net::TcpListener,
gated behind net. axum's own tokio feature would supply it
transitively, but the example names the path directly.
The example reads HOST and PORT (default 127.0.0.1:3000).
PORT=0 lets the E2E tests boot it on an OS-assigned port instead of
racing for one; loopback-by-default came out of Copilot review — this
is code people paste, and it should not expose a routing service on
every interface because someone tried the quickstart. HOST=0.0.0.0
opts in for containers.
The README depends on rustyroute by git, not cargo add rustyroute.
The crate is not on crates.io yet, so the plain command would fail —
also from Copilot review. The "0.1" form is noted for after the
first release.
.gitignore now covers tests/downstream_consumer/target/. The
root /target/ entry is anchored, so reproducing AC3 locally leaves
~300 MB untracked. I hit exactly that.

Acceptance criteria

AC	Status	How it is verified
Example boots; `curl` returns GeoJSON in `[lng, lat]`	✅	`tests/axum_example_e2e.rs` spawns the real binary over a real socket
README snippet compiles under `cargo test --doc`	✅	it compiles and runs; bypass via `rust,ignore` is now blocked
Clean external project gets a valid `Route`	✅	`downstream_consumer` sub-package, plus a fresh out-of-tree crate in QA
13 README groups match `EDGE_GROUPS`, asserted by a test	✅	ordered whole-vector `assert_eq!` against the generated constant

The README's stated quickstart output (16354.1 km over 106 points) is
now asserted too — readme_quickstart_output_matches_reality parses the
numbers out of the README and runs the same route to check them.

Verification

All gates CI runs, run locally, all green:

cargo fmt --all --check                                        ✅
cargo fmt --manifest-path tests/downstream_consumer/… --check  ✅
cargo clippy --all-targets --all-features -- -D warnings       ✅
cargo test --all-features                    ✅ 32 groups, 0 failed
cargo test --doc                             ✅ 3 doctests (2 run, 1 ignored)
RUSTDOCFLAGS="-D warnings" cargo doc --no-deps --all-features  ✅
cargo build --target wasm32-unknown-unknown --no-default-features  ✅
cargo deny check        ✅ advisories, bans, licenses, sources all ok
pre-commit run --all-files                   ✅ no tree drift
cargo package --allow-dirty --list           ✅ ships README + example

Live evidence from a running server (release build):

first position: [5.1927490234375, 43.230220794677734]   ← longitude first
positions: 106   distance_km: 16354.07348421216   resolution: 50
suezCanal blocked: 25047.47884941101 km             ← round the Cape
self-route:        2 positions                      ← RFC 7946 §3.1.4

A fresh crate outside the repo, default features only, running the
README snippet pasted verbatim, prints 16354.1 km over 106 points —
exactly what the README claims.

Notes for the reviewer

Two defects were found by review/QA and fixed here, both worth a look:

A self-route (from == to) returns one coordinate, but a
LineString needs ≥2 — the example would have emitted invalid
GeoJSON. It now pads, as the CLI already does.
Retagging the quickstart ```rust,ignore silently disabled the
doctest gate while every test stayed green. readme_rust_fences_are_exactly_the_two_expected
now blocks it.

Copilot's review then found four more, all fixed in a14db49:

cargo add rustyroute contradicted the status callout two paragraphs
above it — the crate is not published, so the command would fail.
The example bound 0.0.0.0, exposing a pasted quickstart on every
interface. Now loopback by default.
RouteError::NoRoute returned bare text while every other failure
returned {"error": …} JSON, forcing clients to branch on status
before parsing. All errors now leave by one door.
16354.1 km over 106 points was asserted nowhere. Rather than soften
the claim, it is now checked against a real route.

That last fix exposed a weakness in one of my own tests: adding a
#-prefixed TOML comment made the heading-inventory test fail, because
it scanned every line for # and read the comment as an h1. Heading
extraction is now fence-aware.

Every new assertion was proven to fail against the specific defect it
guards before being accepted — including both directions of the new
output check (wrong distance, wrong point count).

Follow-ups (out of scope, not fixed here)

CONTRIBUTING.md still opens with "rustyroute is currently a
repository skeleton… routing algorithms, maritime graph data,
build scripts, and CI workflows are intentionally not yet present" —
all five now exist. The new README links straight to it, so the
contradiction is newly reader-visible.
src/loader.rs documents the OnceLock + Box::leak handle pattern,
which the example mirrors — but Box::leak is redundant there, since
static OnceLock<Graph> already yields &'static Graph.
src/loader.rs:379 says the k-d tree upgrade "is deferred to
ENG-4690", but ENG-4690 shipped as benchmarks (ENG-4690: criterion benchmarks + CodSpeed regression tracking #13).

Checklist

PR title follows the Conventional Commits format above.
All commits are signed off (DCO — git commit -s).
No routing code, vendored data, build script, or CI workflow is
added unless this PR explicitly owns that scope.
LICENSE and NOTICE attribution preserved when those files
were touched.
Community / config files (*.json, *.yml, *.toml, *.md)
parse cleanly with the validators listed in CONTRIBUTING.md.

🤖 Generated with Claude Code

…example Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

…am, and edge-group table Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

…e menai bbox, and its section order Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

… external consumer Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

…hen README drift tests - examples/axum_server.rs: emit graph.resolution_km() instead of a hardcoded 50, so changing the load() call cannot silently report a wrong resolution. - tests/readme_contract.rs: assert the menaiStrait bbox as a whole phrase so an axis swap (lat/lng transposed) fails; add readme_edge_group_table_pass_tags_match_pass_groups so the table's second column is verified against PASS_GROUPS too. - README.md: drop the inaccurate "no build script" claim (rustyroute's own build.rs compiles bundled SQLite and parses ~17 MiB of GeoPackages), note the crate is not yet on crates.io at the point the reader runs cargo add, and say which features the Performance rows need. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

…ckout The peer-review pass edited README.md (build-cost accuracy, crates.io availability note, Performance feature caveats) and re-spliced the axum fence, but a `git checkout README.md` used to revert a deliberate test-discrimination check also discarded those uncommitted edits, and the previous commit captured the reverted file. readme_axum_fence_matches_example_file caught it in the regression run — which is the drift guard doing exactly its job. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

AC1 ("boots; curl returns valid GeoJSON in [lng, lat] order") was the only acceptance criterion with no automated coverage — it was verified by hand during dev and would have rotted on the next handler edit. readme_contract.rs proves the README and the example agree textually; this proves the example actually works. tests/axum_example_e2e.rs spawns the compiled example as a real process and drives it over a real TCP socket with hand-written HTTP/1.1 — no HTTP-client dependency added, since deny.toml gates every new crate. Four tests: coordinate order, RFC 7946 self-route validity, blocked-edge plumbing, and the 400 error contract. The example gains PORT (default 3000) so parallel tests bind OS-assigned ports instead of racing for one; the README fence was re-spliced to match. example_binary() checks the binary's mtime against the source and rebuilds when stale — without that, a filtered `cargo test --test` run picks up a pre-PORT binary and three of four tests lose a race for port 3000, which is exactly how this file failed on its first run. Also gitignore tests/downstream_consumer/target/: the root /target/ entry is anchored, so reproducing AC3 locally leaves ~300 MB untracked. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

…n the E2E harness P2. Retagging the README's library quickstart ```rust,ignore silently disabled AC2: the quickstart stopped being compiled AND run, yet all six drift tests stayed green and CI passed. Verified by doing it — the doctest went from `1 passed; 1 ignored` to `0 passed; 2 ignored`. spec.md section 5 names this as the first forbidden shortcut and plan.md GC16 pinned it by hand, but nothing enforced it. Added readme_rust_fences_are_exactly_the_two_expected, which requires the Rust-family info strings to be exactly ["rust", "rust,no_run"] in order. P3s: - examples: replace graph.resolution_km() with const RESOLUTION_KM. The accessor returns 0 for a Graph::from_bytes handle — the very substitution this file's doc comment recommends for wasm — so a reader following that advice would have served "resolution": 0. - readme_contract: the pass-tag check was a whole-file contains(), so swapping two rows' Source cells passed. Now parses (group, source) rows and checks the pairing. - README: the sample response showed 16354.073, the CLI's {:.3} format; the example serialises a raw f64. Now shows what it actually emits. - e2e: is_fresh only compared against examples/axum_server.rs, so a binary stale w.r.t. src/ was trusted. Now takes the newest mtime across src/, examples/ and Cargo.toml. - e2e: build the Server guard immediately after spawn — Child does not kill on drop, and two panic sites sat before the guard existed. - e2e: stderr is inherited, not piped-and-never-drained; a child panic now reaches the test output instead of filling a pipe. - e2e: memoise example_binary() behind OnceLock so four tests do not each spawn a cargo build against one target dir. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

Re-review confirmed all 8 prior findings fixed and raised 3 more, all real: - e2e: is_fresh() covered src/, examples/ and Cargo.toml but not build.rs, build/** or vendor/**, so the exact staleness class the previous round filed was still reachable — editing PASS_GROUPS reruns build.rs for the lib but not for examples, so a filtered run would assert against a binary whose graph still had the old edge groups. Now seeded with the build inputs too. Verified: touching build/groups.rs triggers a rebuild where it previously did not. - readme_contract: the row-wise rewrite in the last commit made the pass-tag assertion strictly weaker on what a row claims — it checked only for the backticked tag, so "bbox around `suez`" passed. Restored the full "`pass` tag `{tag}`" phrase. Verified it now fails. Also make the read_dir entry error handling consistent with the branch above it: an unreadable entry means rebuild, not skip. - readme_contract: module doc still said five claims / six tests and routed AC2 to lib_rs_compiles_readme_as_doctests — the weak check the new fence test exists to backstop. Updated, with a note saying which of the two is load-bearing so nobody deletes the wrong one. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

Found in exploratory QA: `block=suezCanal,` returned 400 {"error":"unknown edge group: "} — an error naming no group, because the trailing comma produced an empty name. The CLI deliberately does the opposite and its comment names this exact input: dropping empties "covers --block \"\", a trailing comma (--block suezCanal,), and whitespace-only entries — treating those as 'nothing to block' is kinder than reporting an unknown group whose name prints as nothing" (src/bin/rustyroute.rs, run_route). The two surfaces of the crate disagreed, and the HTTP one produced precisely the error the CLI's author had called out. The example now trims and drops empties before edges_for_groups. Regression-tested by blank_block_entries_are_ignored_not_rejected, which asserts each variant routes identically to a bare block=suezCanal rather than merely returning 200. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

readme_edge_group_table_matches_edge_groups_exactly walked the edge-group table inline for column 1, while edge_group_rows() — added in the review round for the row-aware pass-tag check — did the same walk and returned both columns. Two parsers over one table format is what rots: the next person to change the table updates one and leaves the other silently wrong. The test now derives its column-1 vector from edge_group_rows(). 22 lines removed, behaviour identical (both walks trimmed and stripped backticks the same way; edge_group_rows additionally requires a Source cell, which all 13 rows have). Discrimination re-verified after the dedup — a cleanup that quietly made the assertion vacuous would be worse than the duplication it removed. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

codspeed-hq · 2026-07-27T10:21:05Z

Merging this PR will not alter performance

✅ 6 untouched benchmarks

_{Comparing worktree-ENG-4683 (fe2f4f9) with main (c248547)}

Copilot

Pull request overview

Updates the crate’s OSS-facing documentation and adds automated “docs-as-contract” tests to keep the README and quickstarts from drifting away from the actual API and examples.

Changes:

Replaces the placeholder README with an executable quickstart (doctested) plus an axum HTTP server quickstart.
Adds contract tests that pin README sections, edge-group tables, and the axum snippet to code/constants.
Adds an end-to-end integration test that builds/spawns the axum example and validates runtime behavior over a real socket.

Reviewed changes

Copilot reviewed 9 out of 10 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
`README.md`	New OSS README content, including library + axum quickstarts and architecture/feature sections.
`src/lib.rs`	Adds a `#[cfg(doctest)]` README include anchor so README fences are compiled as doctests.
`examples/axum_server.rs`	Adds the axum HTTP server example that the README embeds/pins to.
`tests/readme_contract.rs`	Adds tests that assert README structure and keep README snippets/tables in sync with code.
`tests/axum_example_e2e.rs`	Adds a process-level E2E test that boots the example and validates GeoJSON/output semantics.
`tests/cli_smoke.rs`	Adds an invariant test for single-line `axum`/`tokio` dev-deps (to protect version parsing).
`tests/downstream_consumer/src/lib.rs`	Adds a downstream-consumer helper that runs the README quickstart route via default features.
`tests/downstream_consumer/tests/smoke.rs`	Adds a downstream-consumer smoke test asserting the quickstart produces a plausible route.
`Cargo.toml`	Adds `axum` and `tokio` as dev-dependencies for the example/tests.
`.gitignore`	Ignores the nested downstream-consumer `target/` directory.

Comments suppressed due to low confidence (1)

README.md:66

This dependency snippet implies crates.io (rustyroute = "0.1"), but earlier in the README you state the crate is not yet published. To keep the README self-consistent, use a git/path dependency here until the first crates.io release (or remove the “not yet published” statement).

rustyroute = "0.1"

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot

Pull request overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (2)

README.md:27

The README says the crate is not on crates.io yet, but the installation command shown (cargo add rustyroute) will fail unless the crate is published. Use a git/path dependency form here (or explicitly label this command as “after first release”).

cargo add rustyroute

README.md:66

This dependency snippet uses a crates.io version (rustyroute = "0.1") even though the README states the crate is not published yet. Consider switching this line to a git (or path) dependency to keep the quickstart copy/pasteable today.

rustyroute = "0.1"

- README: `cargo add rustyroute` cannot work — the crate is not on crates.io, which the status callout says two paragraphs earlier. Show the git form that works today, and note the plain form for after the first release. Same for the axum dependency block. - examples: bind 127.0.0.1 by default, not 0.0.0.0. This is code people paste from a README; it should not expose a routing service on every interface because someone tried the quickstart. HOST=0.0.0.0 opts in. - examples: RouteError::NoRoute returned bare text while every other failure returned {"error": ...} JSON, forcing clients to branch on status before parsing. All errors now leave by one door. Asserted by no_route_returns_a_json_error_like_every_other_failure. - README/tests: the claim "This prints 16354.1 km over 106 points" was unasserted anywhere. Rather than soften it to a vague approximation, readme_quickstart_output_matches_reality now parses the numbers out of the README and runs the same route to check them. Verified it fails on both a wrong distance and a wrong point count. Also fixes a weakness the heading test exposed in itself: it scanned every line for `#`, so the new TOML comment inside the dependency fence read as an h1 and failed the inventory. Heading extraction is now fence-aware via prose_lines(); re-verified it still catches a renamed heading. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

Copilot

Pull request overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 1 comment.

The comment showed the listening line as `http://0.0.0.0:34567`, which stopped being true when the example switched to loopback-by-default in a14db49. Misleading exactly when it matters — while debugging a failed port parse. Caught by Copilot; my own change should have updated it. Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

Copilot

Pull request overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

tests/axum_example_e2e.rs:190

start_server inherits the parent process environment, so if a developer has HOST set (common in some shells), the example may bind an unexpected address/hostname and the test can fail/flap. Make the test deterministic by explicitly removing HOST (or setting it to 127.0.0.1) for the spawned child.

    let child = Command::new(example_binary())
        .env("PORT", "0")
        .stdout(Stdio::piped())

…start The Unreleased section already records ENG-4690 and ENG-4692 but had no line for this change. Adds one covering the README rewrite, the runnable axum example, and the drift tests that pin both to the API. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

jimbofreedman · 2026-07-28T02:16:07Z

Handoff verification — `fe2f4f9`

Added the missing CHANGELOG.md [Unreleased] entry for ENG-4683 (the
section already recorded ENG-4690 and ENG-4692 but not this change).
No other change in this push.

Gate	Result
`cargo test` (full suite, local)	✅ exit 0, 0 failed
pre-commit hooks (commit-time, staged files)	✅ all passed
GitHub checks @ `fe2f4f9`	✅ 21/21 pass
DCO	✅ signed off
Review threads	✅ 5/5 resolved, 0 unresolved
`mergeable`	✅ `MERGEABLE`, 0 commits behind `main`

mergeStateStatus is BLOCKED on REVIEW_REQUIRED only — this is a
draft PR awaiting human approval, which is the intended state.

🤖 Generated with Claude Code

Copilot

Pull request overview

Copilot reviewed 10 out of 11 changed files in this pull request and generated 2 comments.

+    let listener = tokio::net::TcpListener::bind(format!("{host}:{port}"))
+        .await
+        .unwrap();
+    println!("listening on http://{}", listener.local_addr().unwrap());
+    axum::serve(listener, app).await.unwrap();


+    let listener = tokio::net::TcpListener::bind(format!("{host}:{port}"))
+        .await
+        .unwrap();
+    println!("listening on http://{}", listener.local_addr().unwrap());
+    axum::serve(listener, app).await.unwrap();


jimbofreedman added 13 commits July 27, 2026 10:11

build(deps): ENG-4683 add axum + tokio dev-dependencies for the HTTP …

ed2c072

…example Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

test(crate): ENG-4683 compile README code fences as doctests

1959b94

Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

docs(crate): ENG-4683 add axum HTTP server example

b414566

Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

docs(crate): ENG-4683 rewrite README with quickstarts, pipeline diagr…

1b73da8

…am, and edge-group table Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

test(crate): ENG-4683 pin README to EDGE_GROUPS, the axum example, th…

d9c194f

…e menai bbox, and its section order Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

test(crate): ENG-4683 assert the README quickstart route works for an…

1acc92e

… external consumer Signed-off-by: Jimbo Freedman <jimbo@spot-ship.com>

Copilot AI review requested due to automatic review settings July 27, 2026 10:19

Copilot started reviewing on behalf of jimbofreedman July 27, 2026 10:20 View session

Copilot AI reviewed Jul 27, 2026

View reviewed changes

Comment thread README.md Outdated

Comment thread examples/axum_server.rs

Comment thread examples/axum_server.rs Outdated

jimbofreedman requested a review from Copilot July 27, 2026 10:25

Copilot started reviewing on behalf of jimbofreedman July 27, 2026 10:25 View session

Copilot AI reviewed Jul 27, 2026

View reviewed changes

Comment thread README.md

jimbofreedman requested a review from Copilot July 27, 2026 10:37

Copilot started reviewing on behalf of jimbofreedman July 27, 2026 10:38 View session

Copilot AI reviewed Jul 27, 2026

View reviewed changes

Comment thread tests/axum_example_e2e.rs Outdated

jimbofreedman requested a review from Copilot July 27, 2026 10:43

Copilot started reviewing on behalf of jimbofreedman July 27, 2026 10:44 View session

Copilot AI reviewed Jul 27, 2026

View reviewed changes

jimbofreedman requested a review from Copilot July 28, 2026 02:39

Copilot started reviewing on behalf of jimbofreedman July 28, 2026 02:40 View session

Copilot AI reviewed Jul 28, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

docs(crate): ENG-4683 README, axum quickstart, and architecture diagram - #20

docs(crate): ENG-4683 README, axum quickstart, and architecture diagram#20
jimbofreedman wants to merge 16 commits into
mainfrom
worktree-ENG-4683

jimbofreedman commented Jul 27, 2026 •

edited

Loading

Uh oh!

codspeed-hq Bot commented Jul 27, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

jimbofreedman commented Jul 28, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

jimbofreedman commented Jul 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Deliberate deviations from the ticket

Acceptance criteria

Verification

Notes for the reviewer

Follow-ups (out of scope, not fixed here)

Checklist

Uh oh!

codspeed-hq Bot commented Jul 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

jimbofreedman commented Jul 28, 2026

Handoff verification — fe2f4f9

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

jimbofreedman commented Jul 27, 2026 •

edited

Loading

codspeed-hq Bot commented Jul 27, 2026 •

edited

Loading

Handoff verification — `fe2f4f9`