Wyrm: memory for AI coding agents

Wyrm gives AI coding agents a memory that lasts between sessions, and stops them repeating mistakes they have already made. It is free, and it runs on your own machine.

Official MCP RegistryListed as lk.ghosts/wyrm (opens in a new tab)
npm install -g wyrm-mcp
Package
wyrm-mcp on npm ↗
Version
9.4.3, released 2026-11-06
MCP Registry
lk.ghosts/wyrm ↗
Needs
Node.js 22 or newer
Licence
Proprietary. Free for local use; paid plans add sync and teams.

What it does

It remembers your project
Decisions, facts about your codebase and what happened in earlier sessions are kept in one file on your machine and handed back to the agent when a new session starts, so you stop explaining the same things again. No account is needed, and by default nothing leaves your machine.
It stops repeated mistakes
When an approach has already failed, Wyrm warns the agent before it tries it again. On the benchmark that ships with Wyrm it caught 16 of 16 repeated mistakes, and wrongly blocked none of 16 new actions.
It works with the tools you use
Claude Code, Claude Desktop, GitHub Copilot, Cursor, Windsurf, Codex, and any other assistant that supports the Model Context Protocol (MCP), the standard these tools use to connect to outside data.

What we have measured

Each figure comes from a benchmark or test suite in the Wyrm repository, and runs offline. The receipts page gives the file, line, commit and date behind every one, and the command that reproduces it.

FigureWhat it measuresCommit, date
100%Repeated mistakes caught, 16 of 16, on the firewall benchmark4a67c24, 2026-07-11
100%New actions let through without a false block, 16 of 164a67c24, 2026-07-11
~0.3 msTime to check an action before it runs (median)6c0b055, 2026-09-14
72.0%Questions where the right memory is in the first ten results (LoCoMo benchmark, the bundled local model)a09f572, 2026-08-14
39.9 → 55.5%Right memory ranked first with NVIDIA NIM embeddings, then with the NIM reranker (optional, 301 questions)3d11168, 2026-10-04
622Jailbreak prompts it was tested against26b244d, 2026-07-17
0Ordinary text wrongly flagged as an attack26b244d, 2026-07-17
3,395Automated tests passing, across 305 suitesc37b496, 2026-09-14

Free on your own machine, with paid plans for sync and teams

Local use is free with no limits and no account: the memory, the mistake firewall and the search, all of it. Paid plans add cloud sync across devices, encrypted snapshots, shared team memory and support. They are arranged by email, not a checkout: write to support@ghosts.lk.

Free

Free

For individual developers

  • Local SQLite storage
  • All 154 tools · 33-verb surface
  • Fleet memory + negative learning
  • Write receipts + trust lanes
  • Local vectors + hybrid recall
  • Cross-project memory
  • Full-text + semantic search
  • Unlimited projects
  • Community support

$29

per month

Pro

For power users

  • Everything in Free
  • Cloud sync across devices
  • Encrypted snapshots (AES-256-GCM)
  • Priority support
  • Analytics

$199

per month

Team

For development teams

  • Everything in Pro
  • Shared team memory
  • Up to 25 seats
  • Admin dashboard

$499

per month

Enterprise

For organizations

  • Everything in Team
  • Unlimited seats
  • SSO / SAML
  • Custom SLA
  • On-premise option

Already on a paid plan? Sign in to your account ↗ Need something the plans do not cover? Contact us.

Local Wyrm stays free. If it is useful to you and you want to support the work, you can buy us a coffee ↗

Recent releases

  1. v9.4.32026-11-06

    cloud sync --all sends only the rows that changed

    Some tables carry no change date, so wyrm cloud sync --all had nothing to filter them on and sent every one of their rows again on every run. The cloud stored each copy as new, and the status command counted them all as pending straight after a clean sync. A local ledger now records what the cloud last stored, so those rows go out only when their content changes. One sync runs on a store at a time, and a flag the command does not know is refused instead of running a full sync.

  2. v9.4.22026-11-03

    Secrets stay out of every column, and a failed push is retried

    One redactor now guards every write, recognises more token formats, and covers text sent to a hosted embedder or reranker. wyrm maintenance redact cleans what is already stored, across every free-text column; run it once after installing. A cloud push stops at the first failed chunk and saves its place only as far as the last unbroken success, and wyrm cloud status shows the rows left behind so the next sync can send them again.

  3. v9.4.12026-10-04

    The word “constructor” no longer stops a memory embedding

    Text containing the word “constructor” or “__proto__” could not be indexed by the bundled local model: the tokenizer read those words as built-in JavaScript objects rather than vocabulary, and the conversion failed. Recall then fell back to keyword search without saying so. Since 9.4.0 indexes private memories on the bundled model, this reached any private memory that mentioned a JavaScript constructor. Other text indexes exactly as before, and the memories that failed are picked up by the next backfill.

All releases
  1. v9.4.02026-10-04

    Private stays private, and memory arrives as data

    A review of 9.3.1 on a store in daily use since June found four problems, and this release fixes all four. On installs using NVIDIA’s hosted models, memories from private projects were sent to NVIDIA to be indexed, which Wyrm’s own egress document says never happens; they are now indexed on the machine. Memory reaches the agent marked as records, not instructions, and an imported rule no longer becomes a ground truth on its own. Backups now run by themselves, wyrm doctor fails when the newest is more than two days old, and servers starting together no longer crash on an upgrade.

  2. v9.3.12026-09-24

    The off switch now switches the renewal off

    The setting WYRM_LICENSE_REFRESH=0 is documented as the switch that stops Wyrm's license calls, but a paid license close to its expiry still renewed itself with the switch set. The check now sits inside the renewal itself, so the switch stops the call it says it stops.

  3. v9.3.02026-09-23

    Signing in keeps a paid license

    Signing in with wyrm login used to replace an installed Pro, Team or Enterprise license with a 30-day free one, without a word. It now keeps a paid license, and installs that were already moved to the free tier fetch theirs back on their own. The same release fixes the setup crash and the hook failures reported from Windows, and wyrm cloud login now tells Wyrm Cloud which machine it is.

  4. v9.2.32026-09-14

    Wyrm keeps its memory honest over time

    A long-term audit of a store in daily use since June found that Wyrm kept everything forever, trusted every truth it was ever told, let its indexes rot, and could silently lose or resurrect data across devices. This release fixes that, and the default tool surface stays at 33 tools and 7,995 tokens.

  5. v9.2.22026-09-14

    Strict arguments and a leaner vault wrapper

    Unknown arguments used to be silently dropped, so a misspelled argument did nothing. Tool calls now reject arguments a tool does not accept, and the wyrm vault exec wrapper that sits in front of every MCP server no longer loads the store, the SDK or the CLI.

  6. v9.2.12026-09-14

    A quality pass across the whole board

    A hands-on audit of 9.2.0 covered the MCP tools, the CLI, the server process, the Claude Code hooks, packaging, and the editor and adapter packages. It found defects that lost no data but made Wyrm harder to trust or to run, and this release fixes them with no schema change and no ranking change.

  7. v9.2.02026-09-13

    NVIDIA retired the NIM models, and nothing said so

    NVIDIA retired both of Wyrm's default hosted NIM retrieval models on 2026-08-25, and Wyrm caught the error and returned nothing, so on installs using the NIM tier new memories stopped being embedded and recall fell back to keyword-only while wyrm doctor still showed a green tick. Both defaults are replaced, and wyrm doctor now checks embedding health.

Common questions

Every figure here traces to the Wyrm repository; the receipts give the file, commit and date behind each one. There is also a page on how Wyrm compares with other AI memory tools.

What is an MCP memory server?

MCP, the Model Context Protocol, is an open standard for connecting AI assistants to outside tools and data. An MCP memory server is one that gives the assistant persistent memory it reads and writes over MCP, so context survives across sessions instead of resetting when the context window clears. Wyrm is an MCP memory server: install it once and any MCP-capable client can use it.

What is Wyrm in one sentence?

Wyrm is a local-first AI memory and intelligence layer, an MCP server that gives coding agents like Claude, Cursor and Copilot persistent memory across sessions and learns from recorded failures to block a known mistake before it repeats.

Who is Wyrm for?

Developers and teams who work in one or more AI coding assistants and want a single persistent memory behind all of them, and people building agent fleets who need run-attributed, accountable memory that many agents can share.

Does Wyrm work with Claude, Cursor and Copilot?

Yes. Wyrm is an MCP server, so it works with any MCP-capable client: Claude Code and Claude Desktop, GitHub Copilot, Cursor, Windsurf, Codex, and others. One memory layer sits behind all of them instead of a separate integration per tool.

Is Wyrm local-first, and where does my data go?

Memory lives in a local SQLite database on your machine. Nothing egresses by default, the free tier needs no account, and it works fully offline. Cloud sync and encrypted snapshots are an explicit opt-in on paid tiers, and any hosted embedding or reranking call is disclosed on every recall.

Is Wyrm free?

Yes. Local use is free with no limits and no account: all 154 tools, the full 33-verb surface, fleet memory, negative learning, write receipts and hybrid recall. Paid tiers add cloud features: Pro is $29 a month, Team $199, Enterprise $499. They are arranged by email, not a checkout: write to support@ghosts.lk.

Is Wyrm open source?

No. Wyrm is proprietary software, free to use on your own machine. Versions 7.2.1 and earlier, published on or before 14 June 2026, were released under an open-source licence, the GNU AGPL version 3, which still applies to those versions only. Paid plans add cloud sync, encrypted snapshots and shared team memory.

How do I install Wyrm?

Run npm install -g wyrm-mcp, then add a server named wyrm pointing at the wyrm-mcp command to your MCP client's config. It is published on npm as wyrm-mcp and on the official MCP Registry as lk.ghosts/wyrm. Setup takes about a minute.

How is Wyrm different from Mem0, Zep and other memory tools?

Two ways. It is local-first and MCP-native, so your memory stays on your machine and plugs into any MCP client without per-app glue. And it does more than store facts: it records what failed and blocks the repeat, links each decision to its reasoning and downstream effects, and injects project ground truths into every context brief. Its published benchmarks are deterministic and reproducible offline, with no cloud LLM in the retrieval path.

What is negative learning?

Wyrm records failures, not just successes. When an agent is about to repeat an approach that already failed, a guard hook surfaces that prior failure and blocks the counter-pattern before the doomed call runs. On the benchmark that ships in the repository it blocks a repeated mistake at 100 percent recall and 100 percent precision, with no false blocks on novel actions, deterministically and in well under a millisecond, no LLM and no cloud.

Does Wyrm need an LLM or the cloud to recall memories?

No. Core recall runs on full-text search fused with local semantic vectors from a bundled embedding model, with no extra LLM call in the hot path, which keeps retrieval fast, cheap and deterministic. Optional NVIDIA NIM embeddings and reranking lift recall further and are an explicit opt-in.

Install Wyrm in one command

npm install -g wyrm-mcp

Then run wyrm-setup to connect your editors. Questions, or a paid plan: support@ghosts.lk. Wyrm is made by Ghost Protocol, a security and software company in Colombo.