mono - fix: authenticated ciphers only, bridge setMany() results, serializer __proto__

[jaredwray]

Please check if the PR fulfills these requirements

• Followed the Contributing and Code of Conduct guidelines.
• Tests for the changes have been added (for bug fixes/features) with 100% code coverage.

What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)

Bug fixes and security hardening, with docs. These are the remaining pre-GA fixes, after jaredwray/keyv#2185 and jaredwray/keyv#2186, one per commit. (A fifth commit added notes on migrating pre-release ciphertext and a sixth reverts it: the encryption adapters haven't been released, so there is nothing to migrate.)

1. encrypt-node: accept only authenticated ciphers and text-safe encodings

KeyvEncryptNode accepted any algorithm Node's getCipherInfo() knows and any BufferEncoding:

• 32 of the algorithm names it accepted failed on the first encrypt(): ECB and other ciphers without an IV, aes-256-xts, chacha20 treated as AEAD, OCB without a tag length, and more.
• utf8 and ascii output never decrypted, and utf16le failed whenever the ciphertext length was odd.
• 77 accepted algorithms (CBC, CFB, CTR, OFB) are unauthenticated: anyone who can write to the store can change values, or read them by watching for decryption errors (a padding oracle).

It now takes an allowlist: aes-128/192/256-gcm, aes-128/192/256-ccm and chacha20-poly1305 (NodeAlgorithm), and base64, base64url or hex (NodeEncoding). Anything else throws when the adapter is created. Every cipher sets a 16-byte tag length, so a truncated tag is rejected. The output format is unchanged, and a test decrypts values written by the previous code. Compatibility between encrypt-node and encrypt-web was never a goal, so the cross-compatibility sections are gone from both READMEs, the encryption overview and encrypt-web's class comment.

2. encrypt-web: accept only AES-GCM

KeyvEncryptWeb also offered AES-CBC, with the same tampering problem. It now accepts only aes-128/192/256-gcm (WebAlgorithm). The lookup uses a Map, so names like constructor no longer pass the check and fail later. The output format is unchanged, and a test decrypts values written by the previous code.

3. keyv: bridge setMany() reports the writes a store rejects

KeyvBridgeAdapter.setMany() returned true for every entry. With a native setMany it dropped the store's result, and without one it ignored each set() result, so keyv.setMany() reported success for writes a legacy store rejected while keyv.set() returned false. It now maps the store's result back to the entries: one boolean per live entry, one for the batch, or nothing (v5 typed setMany as Promise<void>), which still counts as success. Only an explicit false marks a failed write.

4. keyv: the JSON serializer keeps own __proto__ keys

prepare() copied objects into a plain {}, so an own __proto__ key (as JSON.parse creates from a request body) hit the prototype setter and was dropped. v5's @keyv/serialize kept it. The copy now has no prototype. Reading back is unchanged and safe: JSON.parse defines __proto__ as an own property and never sets a prototype.

Verification

• Each fix's new tests fail on main and pass here.
• pnpm test passes in core/keyv (370 tests), encryption/encrypt-node (26) and encryption/encrypt-web (22), with lint clean and every branch of the changed source files covered.
• With the rebuilt keyv, these also pass: bigmap, test-suite, the three compression adapters, both serializers, sqlite, redis (against local Redis), and the website's docs and skill validation.
• tsc --noEmit on encrypt-node reports no errors. encrypt-web's one CryptoKey type error exists on main too, and both packages build.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Wm1wtCGzQJEyYZdekgNVge

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (790eb3c) to head (d302102).

Additional details and impacted files

@@            Coverage Diff            @@
##              main     #2187   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           56        56           
  Lines         5798      5781   -17     
  Branches       997       990    -7     
=========================================
- Hits          5798      5781   -17     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 6b506f22c3

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-03T23:50:14.569525Z	6b506f2	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.