The open alternative to Salesforce, designed for AI.
TypeScript crmcrm-systemcustomergood-first-issuegraphqlhacktoberfestjavascriptmarketingmonoreponestjsopen-sourcepostgresqlreactreactjssalestypescriptweb
chore: rewrite //resolutions to document only current entries (#27217)
Abdullah.3bd7f93cd349
diff --git a/package.json b/package.jsonindex 078ad07580d..391ec1836c7 100644--- a/package.json+++ b/package.json@@ -60,7 +60,7 @@"@mintlify/models/axios": "1.20.0","nx/brace-expansion": "5.0.12"},- "//resolutions": "Each entry is load-bearing: it forces a version OUTSIDE some parent's declared range where no fixed upstream release exists; remove each once its blocker ships. (undici needs NO entry for the September-2026 undici advisory batch, all fixed in 6.28.1 / 7.29.1 / 8.10.2 -- GHSA-rfgv-xxqx-mfg5, GHSA-w293-vg96-wgc3 and GHSA-vp8m-p9jh-q5pm (HIGH) plus eight medium/low -- because every consumer now resolves a fixed version within its own range: @module-federation/dts-plugin since the @module-federation/node ^2.7.21 lift to 2.7.52, which pins @module-federation/enhanced and dts-plugin 2.9.2, and dts-plugin 2.9.2 pins undici 7.29.1 (this retired the former dts-plugin/undici scoped pin, which had chased dts-plugin 2.5.1's exact 7.24.7 through ^7.28.0 / ^7.29.0 / ^7.29.1); miniflare (exact-pinned by wrangler) reaches 7.29.1 through the wrangler ^4.0.0 lift to 4.144.0; e2b's ^7.29.1 -> 7.30.0, node-gyp 12.4.0's ^6.25.0 -> 6.29.0, node-gyp 13's ^8.4.1 and jsdom 30's ^8.9.0 -> 8.11.2.) e2b's OPTIONAL ALIAS dependency undici8 (npm:undici@<exact>) is never touched by the scoped entry above and needs a fixed alias of its own: the @e2b/code-interpreter ^2.6.0 lift to 2.7.1 brought e2b 2.45.0 aliasing [email protected], which retired an earlier alias-form entry but is still inside the September-2026 batch's 8.x range (<8.10.2, incl. the 8.x-only GHSA-vp8m-p9jh-q5pm); e2b's ^2.39.0 range lifts in-range to 2.51.0, which aliases [email protected] (and raises its own undici floor to ^7.29.1), so no alias-form resolution is needed. sharp@npm:0.33.5 + sharp@npm:^0.33.1 -> 0.35.4 are the only sharp entries left for GHSA-f88m-g3jw-g9cj (sharp inherits four libvips CVEs: CVE-2026-33327/33328/35590/35591, vulnerable <0.35.0, fixed 0.35.0) nor for GHSA-rgj7-g3m4-5g8c (sharp <0.35.4 bundles libheif with GHSA-g89c-p67h-r497 / GHSA-2jg2-4ch7-h545, fixed 0.35.4): the former next/sharp ^0.35.3 entry (next's copy ships in scanned images) was dropped once twenty-website's next ^16.2.6 lifted to 16.3.x, which pins sharp ^0.35.3 itself, with @react-email/ui's next forced onto that same version below -- twenty-sdk's direct sharp is bumped to ^0.35.3 in its own package.json, every caret consumer resolves 0.35.4 within its own range, and miniflare's exact pin reached 0.35.4 through the wrangler ^4.0.0 lift to 4.131.x (a [email protected] resolution was briefly considered for it and rejected in favour of that parent bump); @argos-ci/storybook 6.2.3 (twenty-front + twenty-ui) includes the iframe-restoration fix and uses @argos-ci/core 6.7.2 with sharp ^0.35.3, so no argos sharp resolution is needed. The former 6.0.x pin avoided unhandled Monaco worker errors, but retained an iframe-sizing bug that produced oversized screenshots; Monaco workers are now initialized in twenty-front's Storybook preview. sharp 0.33.5 (favicons 7.2.0 + @mintlify/prebuild, website/docs build tooling) sits in both advisory ranges; both parents are at their latest (@mintlify/prebuild 1.0.1255 still exact-pins favicons 7.2.0 and sharp 0.33.5) so no parent lift exists, and it is the last copy keeping Dependabot's root-lockfile sharp alert (#2077) open, so the two descriptors are forced to 0.35.4 -- the exact 0.33.5 pin and favicons' ^0.33.1 range, nothing else. Safe because favicons 7.3.1 itself moved to sharp ^0.35.3 with no API change on the calls favicons 7.2.0 makes; docs/website build only, never in the runtime images. Drop once @mintlify/prebuild pins sharp >=0.35.4 and a favicons that does the same. zapier-platform-core/form-data 4.0.6 -> CVE-2026-12143 / GHSA-hmw2-7cc7-3qxx (form-data CRLF injection via unescaped multipart field names and filenames, vulnerable >=4.0.0 <4.0.6); zapier-platform-core 19.0.0 hard-pins form-data 4.0.5 exact (no caret, still 4.0.5 in its latest) so no parent upgrade carries the fix; scoped to it as the sole 4.0.5 pinner -- every other form-data consumer resolves 4.0.6 naturally via its ^4.0.x range, so the pin dedupes onto it (nx was the other pinner until 22.7.7 moved it to 4.0.6, so its entry was dropped); drop once zapier-platform-core depends on form-data >=4.0.6. (nodemailer needs NO entry: twenty-server and the seed-dependencies declare ^10.0.12 directly, which covers the September-2026 nodemailer advisories fixed across 10.0.2-10.0.9 (GHSA-6vj9-mwq6-2f5v, GHSA-8vvx-rff5-p5rq, GHSA-v53p-9fqp-m79j, GHSA-prgh-xp8r-p3m5, GHSA-g57g-f23g-4646) as well as the earlier 9.1.x ones; v10 ships its own TypeScript declarations, so @types/nodemailer was removed, and imapflow -- which exact-pinned nodemailer through its whole 1.x line -- was lifted to 2.0.2, which drops the nodemailer dependency.) @mintlify/scraping/puppeteer 25.10.0 -> two HIGH extract-zip advisories with NO patched release (extract-zip 2.0.1 is the last release, 2020): GHSA-jmr9-qjv8-65gv (unvalidated symlink path traversal) and GHSA-7pqw-9j4j-h8q3 / CVE-2026-19693. The only consumer is @puppeteer/browsers, which dropped extract-zip for modern-tar in its 3.x line; puppeteer >=25 carries @puppeteer/browsers 3.x, but @mintlify/scraping exact-pins puppeteer 24.3.1 and still does in its latest 4.0.1013, so no parent lift exists. Scoped to @mintlify/scraping (docs tooling, the sole puppeteer consumer in the tree); held at 25.10.0 because 25.11.0 (2026-09-14) is inside the 3d npmMinimalAgeGate. With this pin extract-zip leaves the lockfile entirely. Drop once @mintlify/scraping pins puppeteer >=25. nx/brace-expansion 5.0.12 -> the September-2026 brace-expansion advisories: GHSA-q2hr-2g5m-vwhr (MEDIUM, quadratic-time expansion of the `{a},b}` rewrite), GHSA-6j4f-fj2g-mc7p and GHSA-qhr7-859c-m2p7 (both HIGH, DoS via uncontrolled recursion), all fixed in 5.0.12 on the 5.x line (1.1.21 / 2.1.7 on the older lines). nx exact-pins brace-expansion 5.0.8 and its latest releases still pin vulnerable versions (22.7.12 pins 5.0.8, 23.2.1 pins 5.0.9), so no parent upgrade carries the fix; scoped to nx (its sole exact pinner) -- every other consumer's caret range (^1.1.7 via minimatch 3, ^2.0.x, ^5.0.x) resolves a fixed version on its own; drop once nx pins >=5.0.12. nx/smol-toml 1.7.1 -> GHSA-7w5x-hrqm-74c2 / CVE-2026-85730 (HIGH, DoS via malformed TOML documents, vulnerable <=1.7.0, fixed 1.7.1); nx exact-pins smol-toml 1.6.1 and still does in its latest 23.2.1, so no parent upgrade carries the fix; scoped to nx (its sole consumer); drop once nx pins >=1.7.1. (the former @nestjs/graphql/ws 8.21.0 pin for CVE-2026-48779 was dropped once @nestjs/graphql moved 13.4.2 -> 13.4.5, which pins ws 8.21.3 itself; twenty-server's resolver-scoping patch was rebased onto 13.4.5 for that lift.) @remote-dom/react/@types/react ^19.2.0 -> React type-identity dedup, SCOPED to @remote-dom/react only (every other package resolves @types/react 19 naturally from the workspace ^19.2.0 ranges). @remote-dom/react (transitive via twenty-front-component-renderer) lists @types/react ^18 in its own dependencies and nests its own copy; React 18 and 19 declare ReactNode differently (19 adds bigint + Promise<AwaitedReactNode>, drops ReactFragment's {}), so the two copies are mutually non-assignable and break twenty-front's typecheck (~156 TS2322 errors). Range-aligning our own packages can't fix a third-party's transitive @types pin, hence this single scoped override (only @types/react splits; runtime react/react-dom are peer-deps that converge naturally, and @types/react-dom does not nest a copy). Drop once @remote-dom/react widens @types/react to ^19 (latest 1.2.2 still pins ^18; tracked upstream in Shopify/remote-dom#153). graphql 16.8.1 -> singleton pin held below msw's ^16.12.0 dep and @nestjs/graphql's ^16.11.0 peer; drop after a validated repo-wide bump to latest 16.x; graphql-redis-subscriptions/ioredis 5.10.1 -> TS type-identity dedup: twenty-server passes its ioredis client into RedisPubSub, so this must equal the exact ioredis version pinned by twenty-server and bullmq (bump in lockstep); @types/qs 6.9.16 -> holdback below the 6.9.17 ParsedQs typing break (node-saml wants ^6.9.18); @opentelemetry/api 1.9.1 -> singleton guard for the NoopMeterProvider bug (#20231): ai 6.0.x pins 1.9.0 exact vs @sentry/node ^1.9.1, drop when workspace ai >=6.0.178 AND @scalar/agent-chat moves off ai 6.0.33; chokidar ^3 -> NestJS CLI watch needs fsevents on macOS, removed in chokidar 4/5 (#20316); tmp ^0.2.7 -> CVE, zapier-platform-cli 19 (latest) pins 0.2.5 and inquirer 7/8's external-editor wants ^0.0.33; (the former @mintlify/previewing/tar pin was dropped once the mintlify 4.2.816 lift brought a previewing that pins tar 7.5.21 itself.) @angular-devkit/core 19.2.24 -> picomatch CVE, blocked on @nestjs/cli >11.0.23 fixing the dist/src output regression (repo held at 11.0.16); yeoman-environment 6.0.1 -> CVE, zapier-platform-cli 19 (latest) pins 4.4.3; express/qs 6.16.0 -> two MEDIUM qs advisories fixed in 6.16.0: GHSA-4mjr-xmp4-gh2g / CVE-2026-82417 (DoS via attacker-controlled isBuffer, >=2.2.5 <6.16.0) and GHSA-x5fp-wj9c-mxmx / CVE-2026-82562 (array-limit bypass via bracket-key comma parsing, >=6.14.2 <=6.15.3; lifted from the 6.15.2 pin that covered CVE-2026-8723 / GHSA-q8mj-m7cp-5q26) for old express 4.22.x exact-pinned by @mintlify/previewing (4.22.0 / 4.22.2; its latest 4.0.1371 still pins 4.22.0, while express 4.22.3 itself moved to qs ~6.16.0, so the parent is the blocker); body-parser needs NO entry: the ~1.20.x consumers were lifted in-range to 1.20.8, whose qs ~6.16.0 reaches the fix on its own, and the caret-range qs consumers dedupe onto 6.16.0 naturally; drop once @mintlify/previewing moves off express 4.x or pins >=4.22.3; No postcss resolution remains for the postcss CVEs, latest GHSA-r28c-9q8g-f849 (path traversal in previous source map auto-loading via sourceMappingURL, arbitrary .map file disclosure, vulnerable <=8.5.17, fixed 8.5.18): next 16.3.2 pins 8.5.23 exact itself (the former next/postcss 8.5.23 entry went with the same next lift), the caret consumers (^8.4.38, ^8.4.47, ^8.5.15) dedupe onto 8.5.23 naturally, and @mintlify/common's entry was dropped once the mintlify 4.2.816 lift brought a common that pins 8.5.23 itself; No uuid resolution remains: googleapis-common 8.0.1 (forced via googleapis/googleapis-common) has no uuid dependency so googleapis-common/uuid was dropped, sockjs's entry went when sockjs left the tree and @cypress/request's when the verdaccio lift brought 4.0.1 which has no uuid dependency. @cyntler/react-doc-viewer/ajv 8.20.0 -> CVE, upstream (latest 1.17.1) pins ajv ^7 but never imports it, forcing v8 is safe; */esbuild 0.28.1 -> two esbuild advisories both fixed in 0.28.1: the Deno-module binary-integrity RCE GHSA-gv7w-rqvm-qjhr (vulnerable >=0.17.0 <0.28.1) and the earlier dev-server path-traversal GHSA-g7r4-m6w7-qqqr (Windows, >=0.27.3 <0.28.1). The Deno advisory's range covers every esbuild <0.28.1, so it re-exposed older transitive copies too. Preference is to fix by upgrading the parent, not by resolution -- done where it works: @size-limit/preset-small-lib+size-limit (now ^12.1.0, pin esbuild ^0.28.0) and wrangler (bumped within ^4.0.0 to 4.102.0, which pins esbuild 0.28.1) were bumped and resolve to 0.28.1 on their own, NO resolution needed. The three resolutions below are for parents that pin a vulnerable esbuild OUTSIDE the 0.28.1 range in their latest release: @opennextjs/aws exact-pins 0.25.4 (still 0.25.4 in latest 4.1.0); zapier-platform-cli exact-pins 0.25.8 (still 0.25.8 in latest 19.1.0); @lingui/cli 5.x pins ^0.25.1 -> caps <0.26 (6.x drops esbuild entirely but is a major migration). @react-email/ui (6.9.0 exact-pins 0.28.1), react-email (^0.28.0, no longer down-selected now that 0.28.1 has aged past the npmMinimalAgeGate) and storybook (10.4.6 widened its range to include ^0.28.0) resolve >=0.28.1 on their own, so their entries were dropped. tsx needs NO resolution: its ^4.x ranges resolve to 4.22.x which pins esbuild ~0.28.0 -> 0.28.1 on its own. (tsx had briefly been pinned to 4.21.0 because tsx 4.22's --import tsx/esm loader feeds esbuild-downleveled enums into jest's type-checking ts-node config compiler, yielding a spurious TS7022 in server-integration-test; that is now fixed at the source by dropping --import tsx/esm from the integration jest invocation in twenty-server project.json (it had been swept in by the Storybook 10 upgrade and is not needed there): ts-node alone now compiles the config + globalSetup, so there is no esbuild enum downleveling to trip TS7022 and decorator metadata is still emitted. tsx version is now irrelevant to the tests, so the pin was dropped.) Drop each entry once its parent ships a range that resolves to >=0.28.1 on its own. Our own twenty-client-sdk raises its esbuild floor to ^0.28.1 directly in its package.json instead of via a resolution. googleapis/googleapis-common 8.0.1 -> singleton dedup for the googleapis 173 upgrade. googleapis-common 8.0.2 (published 2026-06-04, AFTER googleapis 173 on 2026-05-28) regressed its google-auth-library dep from a range to exact 10.5.0 (and gaxios to exact 7.1.3), while googleapis itself pulls google-auth-library ^10.2.0 (10.7.0); the two mismatched copies make OAuth2Client's type-identity diverge between the client our provider builds (via googleapis -> 10.7.0) and the one gmail/calendar method options expect (via googleapis-common -> 10.5.0), breaking every gmail/calendar service typecheck (TS2322/TS2769). The parents are already at latest (googleapis 173.0.0, googleapis-common 8.0.2) so we cannot fix by upgrading; instead we pin googleapis' googleapis-common back to 8.0.1, the last version with RANGE deps (google-auth-library ^10.1.0, gaxios ^7.0.0-rc.4) -- and the version googleapis 173 originally shipped against. Ranges resolve to the same hoisted versions googleapis uses, so google-auth-library AND gaxios both collapse to a single copy with no further resolution (a global google-auth-library pin would only dedup one of the two). Scoped to googleapis since it is the sole googleapis-common 8.x consumer. Drop once googleapis-common 8.0.3+ restores range deps or googleapis bumps to a common that does. @verdaccio/config/js-yaml 5.4.2 -> GHSA-r3ph-w7gj-g6xm (MEDIUM, the 5.x line's own maxTotalMergeKeys CPU-bound advisory: empty merge sources are not counted, vulnerable >=5.0.0 <=5.4.0, fixed 5.4.1). @verdaccio/config exact-pins js-yaml 5.2.2 and still does in its latest 8.3.0, so no parent upgrade carries the fix; scoped to @verdaccio/config (the sole 5.x consumer: verdaccio is the local registry the create-app e2e CI publishes to) and held to the latest 5.x patch 5.4.2 rather than the bare 5.4.1 fix -- same major, additive minors. Drop once @verdaccio/config pins js-yaml >=5.4.1. js-yaml@npm:4.3.1 -> 4.3.2 -> GHSA-2883-xcg3-v3hh / CVE-2026-84375 (HIGH, maxTotalMergeKeys does not limit CPU use for empty merge sources, vulnerable >=4.0.0 <4.3.2, fixed 4.3.2; the 3.x line has its own 3.15.2 backport, which @istanbuljs/load-nyc-config's ^3.13.1 reaches on its own). The six @mintlify/* packages (cli, common, prebuild, previewing, scraping, validation) exact-pin js-yaml 4.3.1 and still do in the latest mintlify 4.2.892 / @mintlify/cli 4.0.1495, so no parent upgrade carries the fix; a single descriptor-keyed entry replaces six parent-scoped ones and touches only that exact 4.3.1 pin (every caret consumer dedupes onto 4.3.2 naturally). Their previous six pins had been dropped once the mintlify 4.2.816 lift brought a chain pinning 4.3.1 itself; drop once mintlify pins >=4.3.2. Of the two js-yaml 3.x consumers (both declaring ^3.13.1): @istanbuljs/[email protected] (via babel-plugin-istanbul coverage) is UNPINNED -- the 3.x line shipped backports for every js-yaml advisory (3.15.0 fixes GHSA-h67p-54hq-rp68 + GHSA-52cp-r559-cp3m, 3.15.1 fixes GHSA-5p4m-2wfm-xmqj), so it resolves a clean 3.15.1 within its own range, and as build-only coverage tooling that is enough. [email protected] (via mintlify in twenty-docs) stays FORCED to 4.3.2 (front-matter/js-yaml, lifted in step with the entry above) as a deliberate posture call: 4.x is the maintained js-yaml line, while the 3.x fixes were backports to a legacy major that future advisories may not get, and front-matter parses docs content. Because js-yaml 4.x removed the safeLoad API front-matter's default path calls, the companion patch (front-matter@npm:4.0.2 -> .yarn/patches/front-matter-npm-4.0.2-e1cc0efa69.patch) sets loader = parser.load (safe-by-default in 4.x); pin and patch are a matched pair -- the patch without the pin would select 3.x's UNSAFE full-schema load, so they must only ever be dropped together. The version move stays a separate resolution because a yarn patch only rewrites a package's files, not its resolved dependency graph. Drop both once mintlify moves front-matter off js-yaml 3.x or drops it. @react-email/ui/next 16.3.8 -> Next.js July-2026 security release (CVE-2026-64641 App-Router Server-Actions DoS, CVE-2026-64645 rewrites/redirects SSRF + open redirect, and the sibling CVE-2026-64642..64648), vulnerable >=16.0.0 <16.2.11, fixed 16.2.11; @react-email/ui 6.9.0 hard-pins next 16.2.6 exact (transitive via react-email in twenty-emails). The parent bump was ATTEMPTED and reverted: @react-email/ui + react-email ^6.9.2 do carry next 16.3.0, but that pulls a SECOND next (16.3.0 alongside twenty-website's 16.3.2) plus its dependency subtree into the lockfile -- the scoped pin here forces @react-email/ui's next to the version twenty-website already resolves within ^16.2.6 (16.3.8 since the late-September-2026 CRITICAL GHSA-vcvr-r3jv-pc5j, remote code execution in next/og ImageResponse, >=16.2.0 <16.3.6, fixed 16.3.6; previously 16.3.5 for the September-2026 CRITICAL unauthenticated-RCE advisories GHSA-p293-qw3h-jr36 / CVE-2026-75604 on Windows-hosted servers and GHSA-2xp9-vwfh-vxw4 in the Image Optimization API with AVIF input, both >=16.0.0 <16.3.3, fixed 16.3.3), so it dedupes with ZERO new packages, which is the leaner outcome for a security PR. So @react-email/ui is held at ^6.9.0 with this scoped pin; drop once @react-email/ui itself depends on the next version twenty-website resolves. mintlify axios + adm-zip (twenty-docs) -> bumped mintlify ^4.2.594 -> ^4.2.790 (resolves 4.2.816), which pulls @mintlify/cli >=4.0.1393 -> adm-zip 0.6.0 (fixes CVE-2026-39244 / GHSA-xcpc-8h2w-3j85 for the @mintlify/cli + @mintlify/previewing copies) AND @mintlify/models >=0.0.347 -> axios 1.18.0 (fixes CVE-2026-44494 / SNYK-JS-AXIOS-17111079 axios prototype pollution via config.proxy, vulnerable >=1.15.2 <1.18.0). This one parent bump REPLACED three former scoped resolutions (@mintlify/models/axios, @mintlify/cli/adm-zip, @mintlify/previewing/adm-zip). Blast radius was measured and is contained: ~56 lockfile descriptors change, ALL inside twenty-docs' mintlify subtree (shiki/twoslash, remark/mdx, puppeteer 22->24, and a NESTED type-fest v5) -- the root-hoisted type-fest stays v4 (4.41.0) and every twenty-* workspace package declares type-fest 4.x, so type-fest v5 lives only inside mintlify's own nested tree and touches no workspace typecheck. All typechecks pass (twenty-front / -website / -emails / -ui / -shared reliably; twenty-server passes too modulo the pre-existing flaky tsgo 7.0.0-dev stack-overflow that hits ~1 run in 5 with ZERO type errors), and the mintlify CLI's harmless nested-React 'invalid hook call' warning is byte-identical before and after the bump. nx/axios 1.20.0 + @mintlify/models/axios 1.20.0 -> the September-2026 axios advisory batch, twelve advisories all fixed in 1.20.0, including the HIGH prototype-pollution gadgets GHSA-x97p-jq2g-jp4f / GHSA-m8m8-qj5v-23w3, the HTTP/2 DNS/proxy-control bypass GHSA-3pq3-5fj3-cg6v, the fetch-adapter maxRedirects bypass GHSA-r4gj-5m52-g5wh, the HTTP/2 unhandled-error DoS GHSA-542g-h47m-68v8 and the two ReDoS advisories GHSA-c29m-xwm3-cm6r / GHSA-mghh-pgcx-3jjj (vulnerable ranges vary per advisory, all <1.20.0). Every caret consumer (twenty-server + twenty-sdk ^1.16.0, and transitives) resolves 1.20.0 on its own; these two parents exact-pin a vulnerable axios in their LATEST releases, so no parent upgrade carries the fix: nx pins 1.18.1 through 22.7.12 and 23.2.1 alike, and @mintlify/models pins 1.18.0 through its latest 0.0.363 (mintlify 4.2.964). Both entries are RE-ADDITIONS of previously dropped pins (see below) now that the parents have fallen behind again; drop each once its parent ships axios >=1.20.0. (The former nx/axios pin had been dropped when nx 22.7.7+ moved to axios 1.18.1, >=1.18.0 -- sufficient for the old config.proxy advisory, not for this batch.) adm-zip -> 0.6.1 for the September-2026 adm-zip advisory batch, all vulnerable <=0.6.0 and fixed in 0.6.1: GHSA-8238-w5pm-2374 (HIGH, unhandled error event in async DEFLATE decompression crashes the process), GHSA-rcw4-f5rp-g42v (HIGH, the decompression-bomb guard added for CVE-2026-39244 can be bypassed), GHSA-j5f4-cc29-5x44 (HIGH, extraction preserves SUID/SGID bits from untrusted archives), GHSA-c6fg-446q-cg94 (getDataAsync() bypasses the maxOutputLength guard) and GHSA-p634-w6r4-rjp2 (duplicate entry names resolve inconsistently between getEntry() and extractAllTo()); 0.6.1 also still covers CVE-2026-39244 / GHSA-xcpc-8h2w-3j85 (unbounded Buffer.alloc from a declared uncompressed size, fixed 0.6.0) that the earlier 0.6.0 pins were added for. Three consumers exact-pin a vulnerable adm-zip in their latest releases, so each keeps its OWN scoped pin (not a global one) with its own drop condition: @mintlify/cli and @mintlify/previewing pin 0.6.0 (still 0.6.0 in @mintlify/cli 4.0.1573 / @mintlify/previewing 4.0.1428, mintlify 4.2.970) -- these two entries are RE-ADDITIONS, dropped when the mintlify 4.2.816 lift brought 0.6.0, sufficient then but not for this batch; zapier-platform-cli pins 0.5.16 (still 0.5.16 in its latest 19.1.0, cf. its sibling zapier-platform-core/form-data + zapier-platform-cli/esbuild pins). 0.6.0 -> 0.6.1 is a patch release; the 0.6.0 major's breaking changes (extractEntryTo preserving subdirectories, engines Node >=14) were already absorbed by the zapier pin, and the mintlify copies were already on 0.6.0. (A fourth consumer, @module-federation/dts-plugin, needs no entry since the @module-federation/node lift to 2.7.52 brought dts-plugin 2.9.2, which depends on adm-zip ^0.6.1 itself.) Drop the two mintlify pins once @mintlify/cli and @mintlify/previewing ship adm-zip >=0.6.1; drop the zapier pin once zapier-platform-cli does.",+ "//resolutions": "Each entry forces a version outside some parent's declared range because no parent release carries the fix, or keeps a single copy of a package for type identity where noted; remove each once its drop condition is met. zapier-platform-core/form-data 4.0.6 -> CVE-2026-12143 / GHSA-hmw2-7cc7-3qxx (CRLF injection via unescaped multipart field names and filenames, vulnerable >=4.0.0 <4.0.6); zapier-platform-core exact-pins form-data 4.0.5, still in its latest 19.x. Drop once zapier-platform-core depends on form-data >=4.0.6. @remote-dom/react/@types/react ^19.2.0 -> type identity: @remote-dom/react (via twenty-front-component-renderer) lists @types/react ^18 and nests its own copy; React 18 and 19 declare ReactNode differently, so the two copies are mutually non-assignable and break twenty-front's typecheck. Drop once @remote-dom/react widens @types/react to ^19 (Shopify/remote-dom#153). graphql 16.8.1 -> singleton pin held below msw's ^16.12.0 dependency and @nestjs/graphql's ^16.11.0 peer; drop after a validated repo-wide bump to the latest 16.x. graphql-redis-subscriptions/ioredis 5.10.1 -> type identity: twenty-server passes its ioredis client into RedisPubSub, so this must equal the ioredis version twenty-server and bullmq pin; bump in lockstep with them. @types/qs 6.9.16 -> held below the 6.9.17 ParsedQs typing break (node-saml wants ^6.9.18); drop once twenty-server typechecks against the current @types/qs. @opentelemetry/api 1.9.1 -> singleton guard for the NoopMeterProvider bug (#20231): ai 6.0.33 (via @scalar/agent-chat) pins 1.9.0 exactly while @sentry/node wants ^1.9.1. Drop once @scalar/agent-chat moves off ai 6.0.33. chokidar ^3.6.0 -> NestJS CLI watch mode needs fsevents on macOS, which chokidar 4 and 5 removed (#20316); drop once the NestJS CLI watch works on chokidar 4+. tmp ^0.2.7 -> GHSA-7c78-jf6q-g5cm / CVE-2026-49982 (HIGH, path traversal via a type-confusion bypass of _assertPath, vulnerable >=0.2.6 <0.2.7) and GHSA-ph9p-34f9-6g65 / CVE-2026-44705 (HIGH, path traversal via unsanitized prefix/postfix, vulnerable <0.2.6). zapier-platform-cli 19 (latest 19.1.0) pins tmp 0.2.5 and inquirer 7/8's external-editor wants ^0.0.33. Drop once both resolve tmp >=0.2.7 on their own. @angular-devkit/core 19.2.24 -> brings picomatch 4.0.4 for GHSA-c2c7-rcm5-vvqj / CVE-2026-33671 (HIGH, ReDoS via extglob quantifiers) and GHSA-3v7f-55p6-f55p / CVE-2026-33672 (MEDIUM, method injection in POSIX character classes), both vulnerable >=4.0.0 <4.0.4, reached through @nestjs/cli. Blocked on an @nestjs/cli release above 11.0.23 that fixes its dist/src output regression (the repo holds @nestjs/cli at 11.0.16). yeoman-environment 6.0.1 -> GHSA-vv9j-gjw2-j8wp / CVE-2026-42089 (HIGH, arbitrary package installation, vulnerable >=2.9.0 <6.0.1); zapier-platform-cli 19 (latest 19.1.0) pins 4.4.3. Drop once zapier-platform-cli depends on yeoman-environment >=6.0.1. express/qs 6.16.0 -> GHSA-4mjr-xmp4-gh2g / CVE-2026-82417 (DoS via attacker-controlled isBuffer, >=2.2.5 <6.16.0) and GHSA-x5fp-wj9c-mxmx / CVE-2026-82562 (array-limit bypass via bracket-key comma parsing, >=6.14.2 <=6.15.3) for the express 4.22.0 / 4.22.2 copies @mintlify/previewing exact-pins in its latest releases (express 4.22.3 itself moved to qs ~6.16.0). Drop once @mintlify/previewing pins express >=4.22.3 or moves off 4.x. googleapis/googleapis-common 8.0.1 -> type identity: googleapis-common 8.0.2 pins google-auth-library 10.5.0 and gaxios 7.1.3 exactly while googleapis 173 pulls google-auth-library ^10.2.0, so two OAuth2Client copies diverge and break the gmail and calendar services' typecheck. 8.0.1 is the last release with range dependencies, so both collapse to one copy. Drop once googleapis moves to a googleapis-common with range dependencies. @cyntler/react-doc-viewer/ajv 8.20.0 -> GHSA-2g4f-4pwh-qvx6 / CVE-2025-69873 (MEDIUM, ReDoS when using the $data option, vulnerable >=7.0.0-alpha.0 <8.18.0); @cyntler/react-doc-viewer (latest 1.17.1) declares ajv ^7 but never imports it, so forcing 8.x is safe. Drop once it moves to ajv >=8.18.0 or drops ajv. front-matter/js-yaml 4.3.2 -> front-matter 4.0.2 (via mintlify in twenty-docs) is moved from js-yaml 3.x onto the maintained 4.x line, at 4.3.2 past GHSA-2883-xcg3-v3hh, because it parses docs content and the 3.x line only receives backports. Requires the front-matter@npm:4.0.2 patch below; drop both together, once mintlify moves front-matter off js-yaml 3.x or drops it. @lingui/cli/esbuild 0.28.1 -> GHSA-gv7w-rqvm-qjhr (Deno-module binary-integrity RCE, >=0.17.0 <0.28.1) and GHSA-g7r4-m6w7-qqqr (Windows dev-server path traversal, >=0.27.3 <0.28.1), both fixed in 0.28.1; @lingui/cli 5.x declares esbuild ^0.25.1 (6.x drops esbuild but is a major migration). Drop once @lingui/cli resolves esbuild >=0.28.1 on its own. @opennextjs/aws/esbuild 0.28.1 -> GHSA-gv7w-rqvm-qjhr and GHSA-g7r4-m6w7-qqqr (see @lingui/cli/esbuild); @opennextjs/aws exact-pins esbuild 0.25.4, still in its latest 4.1.6. Drop once @opennextjs/aws pins esbuild >=0.28.1. zapier-platform-cli/esbuild 0.28.1 -> GHSA-gv7w-rqvm-qjhr and GHSA-g7r4-m6w7-qqqr (see @lingui/cli/esbuild); zapier-platform-cli exact-pins esbuild 0.25.8, still in its latest 19.1.0. Drop once zapier-platform-cli pins esbuild >=0.28.1. front-matter@npm:4.0.2 patch -> js-yaml 4.x removed the safeLoad API front-matter's default path calls, so .yarn/patches/front-matter-npm-4.0.2-e1cc0efa69.patch sets loader = parser.load, which is safe by default in 4.x. A matched pair with front-matter/js-yaml: the patch without that pin would select 3.x's unsafe full-schema load, so drop both together. @react-email/ui/next 16.3.8 -> @react-email/ui (via react-email in twenty-emails) exact-pins a vulnerable next: its latest 6.11.0 pins 16.3.3, inside GHSA-vcvr-r3jv-pc5j (remote code execution in next/og ImageResponse, >=16.2.0 <16.3.6). Pinned to the version twenty-website resolves within ^16.2.6 so react-email dedupes onto that single next copy instead of pulling a second next subtree; move it in step with twenty-website's next. Drop once @react-email/ui depends on the next version twenty-website resolves. zapier-platform-cli/adm-zip 0.6.1 -> GHSA-8238-w5pm-2374, GHSA-rcw4-f5rp-g42v and GHSA-j5f4-cc29-5x44 (HIGH) plus GHSA-c6fg-446q-cg94 and GHSA-p634-w6r4-rjp2, all vulnerable <=0.6.0 and fixed in 0.6.1; zapier-platform-cli exact-pins adm-zip 0.5.16, still in its latest 19.1.0. Drop once zapier-platform-cli ships adm-zip >=0.6.1. @mintlify/cli/adm-zip 0.6.1 -> the adm-zip <=0.6.0 advisories (see zapier-platform-cli/adm-zip); @mintlify/cli exact-pins adm-zip 0.6.0 in its latest release. Drop once @mintlify/cli ships adm-zip >=0.6.1. @mintlify/previewing/adm-zip 0.6.1 -> the adm-zip <=0.6.0 advisories (see zapier-platform-cli/adm-zip); @mintlify/previewing exact-pins adm-zip 0.6.0 in its latest release. Drop once @mintlify/previewing ships adm-zip >=0.6.1. nx/smol-toml 1.7.1 -> GHSA-7w5x-hrqm-74c2 / CVE-2026-85730 (DoS via malformed TOML documents, vulnerable <=1.7.0); nx exact-pins 1.6.1, still in 23.2.1. Drop once nx pins >=1.7.1. js-yaml@npm:4.3.1 -> 4.3.2 -> GHSA-2883-xcg3-v3hh / CVE-2026-84375 (maxTotalMergeKeys does not limit CPU use for empty merge sources, vulnerable >=4.0.0 <4.3.2). The six @mintlify/* packages (cli, common, prebuild, previewing, scraping, validation) exact-pin 4.3.1 in their latest releases; descriptor-keyed so only that exact pin moves. Drop once mintlify pins js-yaml >=4.3.2. @verdaccio/config/js-yaml 5.4.2 -> GHSA-r3ph-w7gj-g6xm (the 5.x line's maxTotalMergeKeys advisory, vulnerable >=5.0.0 <=5.4.0, fixed 5.4.1); @verdaccio/config exact-pins 5.2.2, still in its latest 8.3.0 (verdaccio is the local registry the create-app e2e CI publishes to). Held at the latest 5.x patch. Drop once @verdaccio/config pins js-yaml >=5.4.1. sharp@npm:0.33.5 -> 0.35.4 -> GHSA-f88m-g3jw-g9cj (libvips CVEs, vulnerable <0.35.0) and GHSA-rgj7-g3m4-5g8c (bundled libheif, vulnerable <0.35.4); @mintlify/prebuild exact-pins sharp 0.33.5 in its latest release. Descriptor-keyed so only that exact pin moves; docs and website build tooling only, never in the runtime images. Drop once @mintlify/prebuild pins sharp >=0.35.4. sharp@npm:^0.33.1 -> 0.35.4 -> the same sharp advisories (see sharp@npm:0.33.5) for favicons 7.2.0's ^0.33.1 range; @mintlify/prebuild exact-pins favicons 7.2.0 in its latest release. Safe because favicons 7.3.1 itself moved to sharp ^0.35.3 with no API change on the calls favicons 7.2.0 makes. Drop once @mintlify/prebuild pins a favicons that depends on sharp >=0.35.4. @mintlify/scraping/puppeteer 25.10.0 -> extract-zip GHSA-jmr9-qjv8-65gv and GHSA-7pqw-9j4j-h8q3 (symlink path traversal, no patched release); puppeteer >=25 carries @puppeteer/browsers 3.x, which no longer uses extract-zip. @mintlify/scraping 4.0.970 exact-pins puppeteer 24.3.1. Drop once mintlify brings an @mintlify/scraping that no longer depends on puppeteer 24. @ai-sdk/provider@npm:4.0.10 -> 4.0.16 -> type identity, not security: ai 7.0.103 and @ai-sdk/typesafe-ai pin @ai-sdk/provider 4.0.16 exactly while the provider packages twenty-server uses pin 4.0.10 exactly, and two copies make the model types non-assignable. Descriptor-keyed so every 4.0.10 pin collapses onto 4.0.16. Drop once the ai-sdk packages twenty-server uses all pin the same provider version. nx/axios 1.20.0 -> the September-2026 axios advisory batch (twelve advisories, six HIGH, including GHSA-x97p-jq2g-jp4f, GHSA-m8m8-qj5v-23w3 and GHSA-3pq3-5fj3-cg6v), all fixed in 1.20.0; nx exact-pins axios 1.18.1, still in 23.2.1. Drop once nx pins axios >=1.20.0. @mintlify/models/axios 1.20.0 -> the axios advisories fixed in 1.20.0 (see nx/axios); @mintlify/models exact-pins axios 1.18.0 in its latest release. Drop once @mintlify/models pins axios >=1.20.0. nx/brace-expansion 5.0.12 -> GHSA-q2hr-2g5m-vwhr, GHSA-6j4f-fj2g-mc7p and GHSA-qhr7-859c-m2p7 (DoS), fixed in 5.0.12 on the 5.x line; nx exact-pins 5.0.8 (23.2.1 pins 5.0.9). Drop once nx pins brace-expansion >=5.0.12.","version": "0.2.1","nx": {},"scripts": {