Skip to content

build: replace the makefile with mise tasks - #633

Merged
RonsenbergVI merged 5 commits into
mainfrom
build/mise-tasks
Nov 20, 2026
Merged

RonsenbergVI merged 5 commits into
mainfrom
build/mise-tasks

Conversation

@RonsenbergVI

@RonsenbergVI RonsenbergVI commented Nov 20, 2026 •

Copy link
Copy Markdown
Collaborator

Pull Request

Type of change

  • Bug fix
  • New feature
  • Refactor / internal cleanup
  • Docs
  • Other: build tooling and CI

Description of the changes

The Makefile is replaced by mise: mise.toml pins the toolchain (Go 1.26, uv 0.12, pre-commit 4.7.0) and holds every task, named group:target, so make lint-go is now mise run lint:go. CI installs the toolchain with jdx/mise-action, and every step it runs is a mise run, the three releases included (check:tag:go|py|mcp, then release:go|py|mcp), as well as the MCP registry publish, the docs sync to FraiseHQ/www, govulncheck and the example image builds. The benchmarks page moves from the docs into CONTRIBUTING.md.

Motivation for the changes

One file now sets the toolchain locally and in CI, where setup-go read go.mod and setup-uv read pyproject.toml, and a contributor installs all of it with mise install. A CI step that is a task runs the same way on a laptop, with its arguments listed by mise run <task> --help. The docs at docs.getfraise.dev are for users of fraise; how the benchmark gates run and how to accept a regression is for contributors.

Breaking change?

  • No
  • Yes — details below
If yes, fill this in

What breaks:

Before → after:

// before

// after

Migration:

How it was tested

  • mise run lint:all: every pre-commit hook passes; mise run lint passes the hooks no language owns and skips the rest.
  • mise run test:py and mise run test:mcp: both unit suites pass.
  • mise run test:integration (18 passed) and FRAISE_E2E_PORT=9879 mise run test:e2e: both pass, and the e2e container is torn down after.
  • mise run bench:history on sample benchmark outputs: byte-identical to what make bench-history wrote.
  • mise run build:go: the same version stamp as make build-go.
  • mise run check:tag:go|py|mcp <tag>, and with the tag in GITHUB_REF_NAME: a server tag on main and package tags naming the package's version pass; a tag off main and a mismatched version fail.
  • GORELEASER_CURRENT_TAG=v0.3.7 mise run release:go --snapshot --skip=publish,sign,announce: builds every archive, package and the Homebrew formula.
  • mise run build:py and mise run build:mcp: each lands in its package's own dist/.
  • mise run check:vuln: no reachable vulnerabilities. mise run build:examples openai-agents: the image builds.
  • mise run docs:sync against a local bare repository standing in for FraiseHQ/www: removed pages are deleted, files outside contents/docs kept, the commit is authored by fraise-bot[bot] with its id address, and a second run commits nothing.
  • release:registry was not run whole, since it publishes. Its read-only steps were: the server.json read at v0.3.6 is rejected for v0.3.7, and at v0.3.7 the image is pinned to ghcr.io/fraisehq/fraise:0.3.7 and mcp-publisher validate accepts the record.
  • actionlint on every changed workflow. The release, registry and docs-sync jobs run only on a tag, a release or a push to main; every other job runs on this pull request.

Performance regression

Checklist

  • Tests pass locally
  • Added/updated tests for the change (or explained why not): no test code changes, the tasks are exercised by running them as above
  • Docs updated if behavior changed

Notes for reviewers

  • .pre-commit-config.yaml sets language_version: system on golangci-lint-full. pre-commit does not count a go under the home directory, where mise installs it, as the system's, and builds the hook with the latest Go release instead, which golangci-lint v2.12.2 cannot read.
  • mise run lint runs the hooks no language owns, the pull request workflow's slice; lint:go, lint:py and lint:docker run in their own workflows, and lint:all runs every hook, the check before a pull request.
  • Tasks take their inputs as arguments: check:tag:* and release:go:provenance default the tag to GITHUB_REF_NAME, release:registry <tag> --ref <sha>, docs:sync <www> --bot <slug>, build:examples [example]....
  • Tools a single task needs are scoped to it and pinned in mise.lock: goreleaser and cosign (release:go), gh, jq and mcp-publisher (release:registry), govulncheck (check:vuln). They replace goreleaser/goreleaser-action, sigstore/cosign-installer, golang/govulncheck-action and stefanzweifel/git-auto-commit-action. Only mise 2026.11 or later writes task-scoped tools to the lockfile.
  • The MCP registry job now checks out the repository, which on workflow_run is main, so the publish credential only runs main's tasks; the release's server.json is still read as data. mcp-publisher stays pinned by checksum, now in mise.lock.
  • The tag format is no longer checked, in the Docker workflow or in check:tag:*: release-please writes the tags. check:tag:go checks the tag is on main, check:tag:py and check:tag:mcp that it names the package's version.
  • The bench gates take their pytest options directly: mise run bench --bench-baseline=<dir>. The gate workflow passes --bench-always, which now names mise.toml, since it sets the Go version the benchmarks build with.
  • build:py and build:mcp write to the package's own dist/. A bare uv build in this workspace writes to the root's, where uv publish run from the package finds nothing.
  • install-tools, help, clean-all and the Go-only release prep have no task: mise install, mise tasks and clean cover them.

@codecov

codecov Bot commented Nov 20, 2026

Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

@github-actions

github-actions Bot commented Nov 20, 2026 •

Copy link
Copy Markdown

🟢 No regression

Compared with the newest nightly run of main.

internal/graph · BenchmarkSearch
benchmark unit baseline this run vs base
Search/depth0 sec/op 16.2 ms 12.26 ms -24.29% p=0.000
Search/depth1 sec/op 16.53 ms 12.58 ms -23.89% p=0.000
Search/depth2 sec/op 16.36 ms 12.6 ms -23.00% p=0.000
Search/vector sec/op 14.24 ms 10.12 ms -28.98% p=0.000
Search/conversation sec/op 17.01 ms 14.65 ms -13.91% p=0.000
Search/depth0 B/op 6.22 MiB 6.22 MiB +0.00% p=0.027
Search/depth1 B/op 6.399 MiB 6.399 MiB ~ p=0.487
Search/depth2 B/op 6.399 MiB 6.399 MiB +0.00% p=0.017
Search/vector B/op 4.228 MiB 4.228 MiB ~ p=0.706
Search/conversation B/op 6.868 MiB 6.868 MiB ~ p=0.140
Search/depth0 allocs/op 281 281 ~ p=1.000
Search/depth1 allocs/op 450 450 ~ p=1.000
Search/depth2 allocs/op 450 450 ~ p=1.000
Search/vector allocs/op 392 392 ~ p=1.000
Search/conversation allocs/op 645 645 ~ p=1.000
internal/hash · all benchmarks
benchmark unit baseline this run vs base
T1haHash sec/op 10.44 ns 10.57 ns +1.20% p=0.000
XxHash sec/op 16.15 ns 16.04 ns -0.68% p=0.000
T1haHash B/s 4,118,090,000 4,070,180,000 -1.16% p=0.000
XxHash B/s 2,663,070,000 2,680,365,000 +0.65% p=0.000
T1haHash B/op 0 0 ~ p=1.000
XxHash B/op 0 0 ~ p=1.000
T1haHash allocs/op 0 0 ~ p=1.000
XxHash allocs/op 0 0 ~ p=1.000
internal/index · all benchmarks
benchmark unit baseline this run vs base
BTreeIndexSearch sec/op 957.6 µs 900.9 µs -5.92% p=0.000
BTreeIndexSearchWindow sec/op 3.183 ms 2.911 ms -8.55% p=0.000
RPTreeIndexInsert sec/op 72.11 µs 63.34 µs -12.17% p=0.000
RPTreeIndexSearch sec/op 377.3 µs 357.2 µs -5.32% p=0.000
BTreeIndexSearch B/op 1.353 MiB 1.353 MiB ~ p=0.596
BTreeIndexSearchWindow B/op 2.284 MiB 2.284 MiB ~ p=0.675
RPTreeIndexInsert B/op 3.677 KiB 3.652 KiB -0.68% p=0.000
RPTreeIndexSearch B/op 179.1 KiB 179.1 KiB ~ p=0.839
BTreeIndexSearch allocs/op 24 24 ~ p=1.000
BTreeIndexSearchWindow allocs/op 41 41 ~ p=1.000
RPTreeIndexInsert allocs/op 48 48 ~ p=1.000
RPTreeIndexSearch allocs/op 2,115 2,115 ~ p=1.000
internal/query · all benchmarks
benchmark unit baseline this run vs base
RememberCommit/size-100 sec/op 8.53 µs 7.73 µs -9.38% p=0.000
RememberCommit/size-10000 sec/op 8.655 µs 7.792 µs -9.98% p=0.000
RecallCommit/size-1000 sec/op 1.082 ms 1.035 ms -4.37% p=0.000
RecallCommit/size-10000 sec/op 14.65 ms 12.2 ms -16.72% p=0.000
RememberCommit/size-100 B/op 2.486 KiB 2.477 KiB -0.39% p=0.002
RememberCommit/size-10000 B/op 2.504 KiB 2.465 KiB -1.58% p=0.000
RecallCommit/size-1000 B/op 381.5 KiB 381.5 KiB -0.00% p=0.000
RecallCommit/size-10000 B/op 3.53 MiB 3.526 MiB -0.12% p=0.000
RememberCommit/size-100 allocs/op 38 38 ~ p=1.000
RememberCommit/size-10000 allocs/op 38 38 ~ p=1.000
RecallCommit/size-1000 allocs/op 103 103 ~ p=0.474
RecallCommit/size-10000 allocs/op 449 404 -10.02% p=0.000
pkg/server · BenchmarkRetrievalQuality
benchmark unit baseline this run vs base
RetrievalQuality/text/all sec/op 193.4 ms 173.6 ms ~ p=1.000
RetrievalQuality/text/multi-hop sec/op 31.74 ms 27.33 ms ~ p=1.000
RetrievalQuality/text/temporal sec/op 36.44 ms 32.64 ms ~ p=1.000
RetrievalQuality/text/open-domain sec/op 11.76 ms 9.715 ms ~ p=1.000
RetrievalQuality/text/single-hop sec/op 66.5 ms 60.56 ms ~ p=1.000
RetrievalQuality/text/adversarial sec/op 46.72 ms 41.59 ms ~ p=1.000
RetrievalQuality/vectors/all sec/op 265.1 ms 234.9 ms ~ p=1.000
RetrievalQuality/vectors/multi-hop sec/op 44.31 ms 38.37 ms ~ p=1.000
RetrievalQuality/vectors/temporal sec/op 50.93 ms 43.92 ms ~ p=1.000
RetrievalQuality/vectors/open-domain sec/op 15.4 ms 13.38 ms ~ p=1.000
RetrievalQuality/vectors/single-hop sec/op 94.08 ms 83.64 ms ~ p=1.000
RetrievalQuality/vectors/adversarial sec/op 62.59 ms 55.93 ms ~ p=1.000
RetrievalQuality/text/all turn-f1@10 0.1512 0.1512 0.00% exact
RetrievalQuality/text/multi-hop turn-f1@10 0.1059 0.1059 0.00% exact
RetrievalQuality/text/temporal turn-f1@10 0.1671 0.1671 0.00% exact
RetrievalQuality/text/open-domain turn-f1@10 0.09091 0.09091 0.00% exact
RetrievalQuality/text/single-hop turn-f1@10 0.158 0.158 0.00% exact
RetrievalQuality/text/adversarial turn-f1@10 0.1734 0.1734 0.00% exact
RetrievalQuality/vectors/all turn-f1@10 0.1666 0.1666 0.00% exact
RetrievalQuality/vectors/multi-hop turn-f1@10 0.1506 0.1506 0.00% exact
RetrievalQuality/vectors/temporal turn-f1@10 0.1818 0.1818 0.00% exact
RetrievalQuality/vectors/open-domain turn-f1@10 0.1543 0.1543 0.00% exact
RetrievalQuality/vectors/single-hop turn-f1@10 0.1606 0.1606 0.00% exact
RetrievalQuality/vectors/adversarial turn-f1@10 0.1773 0.1773 0.00% exact
RetrievalQuality/text/all turn-p@1 0.3299 0.3299 0.00% exact
RetrievalQuality/text/multi-hop turn-p@1 0.125 0.125 0.00% exact
RetrievalQuality/text/temporal turn-p@1 0.5946 0.5946 0.00% exact
RetrievalQuality/text/open-domain turn-p@1 0.09091 0.09091 0.00% exact
RetrievalQuality/text/single-hop turn-p@1 0.3286 0.3286 0.00% exact
RetrievalQuality/text/adversarial turn-p@1 0.3191 0.3191 0.00% exact
RetrievalQuality/vectors/all turn-p@1 0.3858 0.3858 0.00% exact
RetrievalQuality/vectors/multi-hop turn-p@1 0.25 0.25 0.00% exact
RetrievalQuality/vectors/temporal turn-p@1 0.6486 0.6486 0.00% exact
RetrievalQuality/vectors/open-domain turn-p@1 0.09091 0.09091 0.00% exact
RetrievalQuality/vectors/single-hop turn-p@1 0.3857 0.3857 0.00% exact
RetrievalQuality/vectors/adversarial turn-p@1 0.3404 0.3404 0.00% exact
RetrievalQuality/text/all turn-recall@10 0.7652 0.7652 0.00% exact
RetrievalQuality/text/multi-hop turn-recall@10 0.3047 0.3047 0.00% exact
RetrievalQuality/text/temporal turn-recall@10 0.9189 0.9189 0.00% exact
RetrievalQuality/text/open-domain turn-recall@10 0.2727 0.2727 0.00% exact
RetrievalQuality/text/single-hop turn-recall@10 0.8571 0.8571 0.00% exact
RetrievalQuality/text/adversarial turn-recall@10 0.9362 0.9362 0.00% exact
RetrievalQuality/vectors/all turn-recall@10 0.8236 0.8236 0.00% exact
RetrievalQuality/vectors/multi-hop turn-recall@10 0.4141 0.4141 0.00% exact
RetrievalQuality/vectors/temporal turn-recall@10 1 1 0.00% exact
RetrievalQuality/vectors/open-domain turn-recall@10 0.5455 0.5455 0.00% exact
RetrievalQuality/vectors/single-hop turn-recall@10 0.8714 0.8714 0.00% exact
RetrievalQuality/vectors/adversarial turn-recall@10 0.9574 0.9574 0.00% exact
RetrievalQuality/text/all B/op 65.01 MiB 65.27 MiB ~ p=1.000
RetrievalQuality/text/multi-hop B/op 10.95 MiB 10.82 MiB ~ p=1.000
RetrievalQuality/text/temporal B/op 12.4 MiB 12.57 MiB ~ p=1.000
RetrievalQuality/text/open-domain B/op 3.823 MiB 3.724 MiB ~ p=1.000
RetrievalQuality/text/single-hop B/op 22.59 MiB 22.69 MiB ~ p=1.000
RetrievalQuality/text/adversarial B/op 15.32 MiB 15.48 MiB ~ p=1.000
RetrievalQuality/vectors/all B/op 91.77 MiB 91.62 MiB ~ p=1.000
RetrievalQuality/vectors/multi-hop B/op 15.15 MiB 15.13 MiB ~ p=1.000
RetrievalQuality/vectors/temporal B/op 17.29 MiB 17.47 MiB ~ p=1.000
RetrievalQuality/vectors/open-domain B/op 5.427 MiB 5.31 MiB ~ p=1.000
RetrievalQuality/vectors/single-hop B/op 31.88 MiB 31.87 MiB ~ p=1.000
RetrievalQuality/vectors/adversarial B/op 21.8 MiB 21.72 MiB ~ p=1.000
RetrievalQuality/text/all allocs/op 108,540 108,379 ~ p=1.000
RetrievalQuality/text/multi-hop allocs/op 16,889 16,861 ~ p=1.000
RetrievalQuality/text/temporal allocs/op 19,402 19,450 ~ p=1.000
RetrievalQuality/text/open-domain allocs/op 6,156 6,108 ~ p=1.000
RetrievalQuality/text/single-hop allocs/op 37,564 37,605 ~ p=1.000
RetrievalQuality/text/adversarial allocs/op 25,126 25,217 ~ p=1.000
RetrievalQuality/vectors/all allocs/op 205,314 205,002 ~ p=1.000
RetrievalQuality/vectors/multi-hop allocs/op 33,161 33,143 ~ p=1.000
RetrievalQuality/vectors/temporal allocs/op 38,227 38,263 ~ p=1.000
RetrievalQuality/vectors/open-domain allocs/op 11,855 11,773 ~ p=1.000
RetrievalQuality/vectors/single-hop allocs/op 70,773 70,665 ~ p=1.000
RetrievalQuality/vectors/adversarial allocs/op 48,312 48,284 ~ p=1.000

@RonsenbergVI
RonsenbergVI merged commit 9c693be into main Nov 20, 2026
54 of 58 checks passed
@RonsenbergVI
RonsenbergVI deleted the build/mise-tasks branch November 20, 2026 16:05
@fraise-bot fraise-bot Bot mentioned this pull request Nov 17, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant