{"id":"CVE-2026-55575","aliases":["GHSA-g357-x5c3-c72p"],"title":"LiquidJS: `pop` filter bypasses `memoryLimit` accounting that its array-filter siblings enforce","summary":"LiquidJS: `pop` filter bypasses `memoryLimit` accounting that its array-filter siblings enforce","severity":"high","cwe":["CWE-770"],"vendor":"liquidjs","product":"liquidjs","ecosystem":"npm","affected":["liquidjs <= 10.27.0"],"patched":["liquidjs 10.27.1"],"published":"2026-07-24","updated":"2026-07-24","source":"GHSA","sourceUrl":"https://github.com/advisories/GHSA-g357-x5c3-c72p","references":[{"url":"https://github.com/harttle/liquidjs/security/advisories/GHSA-g357-x5c3-c72p"},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-55575"},{"url":"https://github.com/harttle/liquidjs/pull/907"},{"url":"https://github.com/harttle/liquidjs/commit/8a0c74a7fcb1671aa1dcb71ec82ba0602dc90d04"},{"url":"https://github.com/harttle/liquidjs/releases/tag/v10.27.1"},{"url":"https://github.com/advisories/GHSA-g357-x5c3-c72p"}],"tags":["ghsa","npm"],"epss":0.00524,"epssPercentile":0.43404,"ingestedAt":"2026-07-24T14:29:29.373Z","slug":"CVE-2026-55575","body":"## Overview\n\n# `pop` filter bypasses `memoryLimit` accounting that its array-filter siblings enforce\n\n**CWE**: CWE-770 (Allocation of Resources Without Limits or Throttling) — sibling class of GHSA-8xx9-69p8-7jp3 and GHSA-2546-xv4c-mc8g, applied to `memoryLimit` instead of `renderLimit`\n\n## Summary\n\nThe `pop` array filter at `src/filters/array.ts:91-95` allocates a full clone of its input array via `[...toArray(v)]` but does **not** call `this.context.memoryLimit.use(...)` the way every other array-clone filter in the same file does (`shift`, `unshift`, `compact`, `concat`, `reverse`, `sample`, `slice`, `map`, `sortBy`, `where`, `group_by`, `uniq`). This silently disables the `memoryLimit` budget for `{{ huge_array | pop }}`, letting a template render allocate an O(N) clone of an attacker-influenced array regardless of how strictly `memoryLimit` is set.\n\n## Affected\n\n- liquidjs ≥ all versions that ship the current `pop` filter implementation (verified `10.27.0`, HEAD `a8fd734b5`)\n- Deployments where any template uses `{{ arr | pop }}` on an array whose length is influenced by untrusted input (typical multi-tenant context arrays: orders, log lines, catalog entries, user lists, etc.)\n\n## Vulnerability details\n\n### Code\n\n`src/filters/array.ts:91-95`:\n\n```ts\nexport function pop<T> (v: T[]): T[] {\n  const clone = [...toArray(v)]   // O(N) allocation — not charged to memoryLimit\n  clone.pop()\n  return clone\n}\n```\n\nNote: the function signature does not even declare `this: FilterImpl`, so it has no typed access to `this.context.memoryLimit` at the type level — a visual tell that the author skipped the limit-accounting boilerplate the surrounding filters use.\n\nCompare with `shift` (`src/filters/array.ts:97-103`), which is functionally identical except for the array-end operated on:\n\n```ts\nexport function shift<T> (this: FilterImpl, v: T[]): T[] {\n  const array = toArray(v)\n  this.context.memoryLimit.use(array.length)   // ← guard present\n  const clone = [...array]\n  clone.shift()\n  return clone\n}\n```\n\nAnd `unshift`, `compact`, `concat`, `reverse`, `sample`, `slice`, `map`, `sortBy`, `where`, `group_by`, `uniq` — all of which also charge `memoryLimit.use(array.length)` (or `lhs.length + rhs.length` etc.) before allocating their working buffer.\n\nThe asymmetry confirms `pop` is an accidental omission, not by design.\n\n### Why the bypass matters\n\n`memoryLimit` is the documented control for bounding the memory a single `render()` call may allocate (`docs/source/tutorials/dos.md`). Every array-output filter in `src/filters/array.ts` other than `pop` deducts its working set from the limit, so a render that does `{{ huge | shift }}` with `memoryLimit: 100` and `huge.length === 5_000_000` correctly throws `memory alloc limit exceeded`. The identical `{{ huge | pop }}` does **not** throw — the allocation proceeds, and the only ceiling is the Node process's heap.\n\n## Proof of concept\n\n```js\nconst { Liquid } = require('liquidjs');\n\nconst l    = new Liquid({ memoryLimit: 100 });        // 100-unit budget\nconst huge = Array(5_000_000).fill('x');              // 5M-element context array\n\n(async () => {\n  try { await l.parseAndRender('{{ a | shift | size }}', { a: huge }); }\n  catch (e) { console.log('shift:   ' + e.message); }      // expected: memory alloc limit exceeded\n\n  try { await l.parseAndRender('{{ a | unshift: 0 | size }}', { a: huge }); }\n  catch (e) { console.log('unshift: ' + e.message); }      // expected: memory alloc limit exceeded\n\n  const out = await l.parseAndRender('{{ a | pop | size }}', { a: huge });\n  console.log('pop:     OK, size=' + out);                  // size=4999999 — allocation succeeded\n})();\n```\n\nObserved (against `dist/liquid.node.js` at `a8fd734b5`):\n\n```\nshift:   memory alloc limit exceeded, line:1, col:1\nunshift: memory alloc limit exceeded, line:1, col:1\npop:     OK, size=4999999\n```\n\n## Impact\n\n- **`memoryLimit` does not bound `pop` allocations.** Any template that can reach `{{ <untrusted-sized array> | pop }}` allocates an O(N) clone outside the budget.\n- **Realistic attack surface**: when a server passes an attacker-influenced large array to the template context (search results, paginated lists, batch-export pages) and the template uses `| pop` anywhere on it, a single render can allocate hundreds of MB of array slots that the operator believed `memoryLimit` had ruled out.\n- **Concurrent amplification**: N parallel requests each allocate their own unguarded clone — the practical ceiling is the Node process heap, after which the host runs `oom-kill`. This is the same outcome the renderLimit-empty-body advisories (GHSA-8xx9-69p8-7jp3 / GHSA-2546-xv4c-mc8g) prevented for CPU; this report prevents it for memory.\n\nSeverity is configuration-dependent (requires `memoryLimit` to be set, plus a template that uses `pop`, plus attacker-influenced array length). For deployments that rely on `memoryLimit` as a DoS guard, this is a real bypass of that guard.\n\n## Workaround for users\n\nUntil a fix lands, deployments relying on `memoryLimit` should either:\n\n- Avoid `| pop` in templates whose inputs include untrusted-length arrays. Use `| slice: 0, arr.size | minus: 1` or equivalent guarded alternatives.\n- Register a wrapping `pop` filter that does the accounting:\n\n  ```js\n  liquid.registerFilter('pop', function (v) {\n    const arr = Array.from(v ?? []);\n    this.context.memoryLimit.use(arr.length);\n    arr.pop();\n    return arr;\n  });\n  ```\n\n## Suggested fix\n\nOne-line addition mirroring `shift`:\n\n```ts\nexport function pop<T> (this: FilterImpl, v: T[]): T[] {\n  const array = toArray(v)\n  this.context.memoryLimit.use(array.length)   // ← add this line, and add `this: FilterImpl`\n  const clone = [...array]\n  clone.pop()\n  return clone\n}\n```\n\nNo API or behavior change for callers within budget; rejects out-of-budget calls with the standard `memory alloc limit exceeded` exception the sibling filters already throw.\n\n## Affected packages\n\n- `liquidjs <= 10.27.0`\n\n## Remediation\n\nUpgrade to a patched release:\n\n- `liquidjs 10.27.1`","depth":"twilight","depthScore":41,"depthScoreParts":{"impact":41.3,"likelihood":0.1,"exploitation":0,"ransomware":0},"changes":[]}