Implement atomic cache primitives and compatibility roadmap

[konard]

Summary

This turns the competitor roadmap into working code by completing its first independently testable implementation phase:

• add conditional add and replace operations
• add TTL refresh through touch, getex, and getAndTouch
• add atomic fetch/delete through getdel and getAndDelete
• add numeric increment/decrement operations and incr/decr aliases
• add string append and prepend operations
• serialize read-modify-write operations across cache instances in the same JavaScript process
• write entries through atomic file replacement and refresh single-file stores from disk
• expose the new API in TypeScript declarations, README documentation, and release notes
• maintain an explicit Redis, Memcached, and Dragonfly compatibility matrix and phased roadmap

Reproduction and regression coverage

Before this change the documented Phase 1 methods did not exist, and two cache instances could not safely express conditional or read-modify-write operations. The new cross-mode regression tests exercise every operation in both folder and single-file storage, including TTL preservation, expired entries, invalid value types, concurrent counters, and competing conditional adds.

The implementation guarantees atomic read-modify-write behavior across LinoCache instances in one JavaScript process. Cross-process coordination, wire protocols, collection commands, persistence, replication, and cluster behavior remain explicitly tracked as later roadmap phases; this PR does not claim Redis, Memcached, or Dragonfly wire compatibility.

Verification

• npm test — 46 passed
• bun test — 46 passed
• deno test --allow-read --allow-write — 46 passed
• npm run check — ESLint, Prettier, and duplication checks passed
• node scripts/validate-changeset.mjs — passed

Addresses #3.

Fixes #3

[konard]

Working session summary

Implemented and finalized PR #4.

Key changes:

• Added a Redis, Memcached, and Dragonfly compatibility matrix.
• Distinguished semantic, client API, and wire-protocol compatibility.
• Documented current gaps and a six-phase implementation roadmap.
• Linked the matrix from the README.
• Updated the release changeset.
• Marked the PR ready for review.

Verification:

• 30 local tests passed.
• Lint, Prettier, duplication, and changeset validation passed.
• Full CI passed across Node.js, Bun, and Deno on Linux, macOS, and Windows.
• Working tree is clean.

---

This summary was automatically extracted from the AI working session output.

[konard]

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost estimation:

• Model: GPT-5.6 Sol
• Provider: OpenAI
• Public pricing estimate: $2.467523

📊 Context and tokens usage:

• 68.1K / 200K (34%) input tokens, 7.8K / 128K (6%) output tokens

Total: (68.1K + 1.4M cached) input tokens, 7.8K output tokens, $2.467523 cost

🤖 Models used:

• Tool: OpenAI Codex
• Requested: gpt-5.6-sol
• Thinking level: medium (~15999 tokens)
• Model: GPT-5.6 Sol (gpt-5.6-sol)

📎 Log file uploaded as Gist (1069KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

✅ Ready to merge

This pull request is now ready to be merged:

• All CI checks have passed
• No merge conflicts
• No pending changes

---

Monitored by hive-mind with --auto-restart-until-mergeable flag

[konard]

Not only document, but actually implement in this pull request.

[konard]

🤖 AI Work Session Started

Starting automated work session at 2026-07-27T05:31:26.608Z

The PR has been converted to draft mode while work is in progress.

This comment marks the beginning of an AI work session. Please wait for the session to finish, and provide your feedback.

[konard]

I’m implementing the first independently testable roadmap phase in this PR: add/replace, touch, numeric mutation, append/prepend, get-and-delete/get-and-touch, with tests in both storage modes and updated API/type documentation. The issue says “all features,” while phases 2–6 cover separate storage, protocol-server, and distributed-system projects. Please confirm whether you expect those later phases in this same PR as well, and if so which protocol/distributed compatibility contract should be authoritative.

[konard]

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost estimation:

• Model: GPT-5.6 Sol
• Provider: OpenAI
• Public pricing estimate: $10.724574

📊 Context and tokens usage:

• 271.0K / 200K (136%) input tokens, 35.1K / 128K (27%) output tokens

Total: (271.0K + 6.4M cached) input tokens, 35.1K output tokens, $10.724574 cost

🤖 Models used:

• Tool: OpenAI Codex
• Requested: gpt-5.6-sol
• Thinking level: xhigh (~31999 tokens)
• Model: GPT 5.6 Sol (gpt-5.6-sol)

📎 Log file uploaded as Gist (2549KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

🎉 Auto-merged

This pull request has been automatically merged by hive-mind.

• All CI checks have passed

---

Auto-merged by hive-mind with --auto-merge flag