Inventory

About Inventory

Inventory is a local-only search index for conversations from Cursor, Claude Code, Zed, Codex, and Kiro. It runs on your machine with no signups, no cloud sync, and charges a one-time fee. The tool indexes both historical and new conversations so you can search across multiple AI agents and IDEs from a single bar.

Review

Inventory launched this week as a response to a specific problem: AI tool users accumulate conversations across several apps and have no way to search them. The maker, Neil Shah, built it because he kept losing track of which tool held a past session. The tool's core constraint - local-only storage with no cloud component - is a deliberate design choice, not a limitation waiting to be fixed.

Key Features

• Indexes all historical conversations from Cursor, Claude Code, Zed, Codex, and Kiro at install time, not just new ones going forward.
• Runs entirely on the local machine with no account creation, no data upload, and no subscription required.
• Version 1.0.2 ships with SQLCipher encryption where the encryption key is generated once per machine and stored in the macOS Keychain, not in the database file.
• If a tool's local storage format changes in an update, Inventory's index for that tool stops working until a fix ships. The current release includes a self-repair mechanism that re-runs on next launch rather than recording a failed parse as a completed read.
• One search bar searches across all indexed tools simultaneously.

Pricing and Value

Inventory uses a one-time fee model. There is no subscription, no cloud tier, and no free trial mentioned in the launch materials. A 30% launch discount is active for the initial release. The pricing page on the tool's website handles payment collection, but specific dollar amounts are not stated in the launch discussion.

Pros

• Indexes conversations you already have - it doesn't start collecting data only from the install date.
• No data leaves your machine, which matters for users who keep client context or proprietary code in their AI tool conversations.
• Encryption is handled via SQLCipher with the key in macOS Keychain, so the index file on disk is opaque to other processes or copied backups.
• If the encryption key is lost or rotated, a full reindex from the original tool files recovers everything without data loss.
• One-time payment with no recurring costs.

Cons

• Currently scoped to a single user's machine - it does not support team-wide search across multiple accounts or shared conversations.
• If a supported tool changes its local storage format, Inventory's index for that tool stops working until the developer ships a patch. The self-repair mechanism helps but doesn't prevent the break.
• Not well suited for users who need cloud sync, cross-device search, or who work on Windows or Linux - the encryption and Keychain integration is macOS-specific.

Inventory is built for individual developers who run multiple AI coding tools and want to search past conversations without uploading anything to a server. If you use Cursor, Claude Code, or similar tools and have ever spent twenty minutes hunting for a past session, the tool solves that exact workflow. It is not a team product, nor does it try to be one in its current form.