# Fix list: Memory (MCP reference server) From Anchor Terminal's listing at https://www.anchorterminal.com/tools/memory-reference-server, the October 2026 research run, assessed 1 October 2026. Grade C, 54.4 out of 100. This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public. For a coding agent working on Memory (MCP reference server): work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published. ## 1. Security & auth, 32 out of 100, up to 11.9 more on the total Why it scored 32: No credentials, nothing to scope, so the middle band (20). No read-only mode. The three delete tools run without confirmation and annotations are the only signal a host gets. The file path is whatever `MEMORY_FILE_PATH` says (5). Everything stored comes back verbatim in later sessions, so an instruction an agent copied from a web page or document persists as memory. The README has no guidance on this (2). No log of who changed what. Resource notifications say only that the graph changed (0). SECURITY.md says the repository isn't eligible for vulnerability reports. No advisories exist for this server, and the repository has published them for others (5). The checklist (https://www.anchorterminal.com/benchmark/#checklist-security): - 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option. - 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions. - 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10. - 0 to 15, audit logs or per-call visibility for the operator. - 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public. Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing. ## 2. Reliability, 52 out of 100, up to 9.6 more on the total Why it scored 52: Scored as a local stdio package. Official npm package @modelcontextprotocol/server-memory and the mcp/memory Docker image, with no Node version stated (15). The monorepo's TypeScript workflow runs Vitest on push and pull request, and the memory server has 7 test files. We couldn't confirm the current head passes, so 20 of 25. Eleven memory fixes merged on 2 and 3 September 2026, among them the concurrent-write race reported in #1819, validation of the file on load and rejection of dangling relations, but none has shipped. The published 2026.8.31 writes atomically and still lets two calls in one turn overwrite each other's changes (12). Date versions with release notes that list package names only (5). The repository calls its servers reference implementations, not production-ready (0). The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability): Hosted APIs, MCP servers, models and platforms. - 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own). - 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so. - 15, rate limits documented with numbers. - 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved. - 10, an SLA published for any paid tier. - 10, the surface agents use is generally available, not beta or preview. Local packages, SDKs, frameworks and stdio MCP servers. - 20, installs from an official package with supported runtimes stated. - 25, a public CI and test suite, passing on the default branch. - 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered). - 15, semver discipline and breaking changes called out in a changelog. - 15, version 1.0 or later, or declared stable. Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors. ## 3. Schema & documentation, 60 out of 100, up to 6.5 more on the total Why it scored 60: All nine tools take typed JSON Schema built from zod and declare output schemas (25). No llms.txt. The README is Markdown on GitHub with JSON examples (5). Descriptions are one line of purpose ("Read the entire knowledge graph"). Only `create_relations` adds guidance ("Relations should be in active voice"), and none says when to prefer `search_nodes` or `open_nodes` over `read_graph` (6). Required fields are marked and every field has a description, but arrays have no minimum or maximum, `entityType` and `relationType` are free strings, and the 2,048-character cap on `search_nodes` sits on main, not in the release (7). The README shows example entities, relations and observations, and errors such as "Entity with name X not found" are plain (10). Dated npm versions with no changelog (7). The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema): APIs and MCP servers. - 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool). - 10, llms.txt or Markdown docs served for agents. - 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference. - 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs. - 0 to 15, examples and documented error responses. - 15, versioning and a public changelog. Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference. ## 4. Agent ergonomics, 67 out of 100, up to 5.4 more on the total Why it scored 67: Nine tools with 560 characters of descriptions in total, but the entity and relation schemas repeat in every output schema, so tools/list runs to about 10,700 characters or roughly 2,700 tokens by our count (22). No pagination, limit or field selection anywhere. `read_graph` returns the whole graph and `search_nodes` returns every substring match (5). Missing entities raise a named error. In the published version, deletes report success whether or not anything matched and relations to missing entities are accepted silently. Both are fixed on main and unreleased (10). All nine tools carry `readOnlyHint`, `destructiveHint` and `idempotentHint`, the three deletes marked destructive (20). At most one required parameter per tool. Node only, plus Docker (10). The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics): - 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries). - 20, pagination, filtering and output-size controls. - 20, actionable, documented error responses, codes and messages an agent can recover from. - 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations. - 15, sensible defaults, few required parameters, and official SDKs in at least two languages. Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs. ## 5. Payments & pricing, 60 out of 100, up to 5 more on the total Why it scored 60: Free, self-hosted, nothing to buy, so 20 + 20 + 20. No payment protocol (0). The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments): The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/). - 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which. - 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login. - 20, a free tier or trial that doesn't need a card. - 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API). Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied. Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol. ## 6. Maintenance & community, 43 out of 100, up to 5 more on the total Why it scored 43: 2026.8.31 on 31 August, 31 days before the run date (20). Only two releases changed this server since 3 July, 2026.7.4 and 2026.8.31 (0). Eleven memory pull requests merged on 2 and 3 September, including the fix for #1819, a race reported in 2025, but a month later none is released. GitHub's robots rules blocked issue search, so reply times are unread (15). Not in the official MCP registry. A lookup for io.github.modelcontextprotocol/server-memory returns 404 although package.json carries that `mcpName` (0). Dependabot, OIDC trusted publishing and SDK ^1.30.0, but the zod runtime dependency is only declared on main (8). The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance): - 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older. - 20, at least three releases or dated changelog entries in the last 90 days. - 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15. - 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models). - 10, package health, current dependencies and CI. Models are read for deprecation notice periods and model churn rather than release counts. ## 7. Transparency & trust, 74 out of 100, up to 2.3 more on the total Made of editorial 74, provenance 74. Why it scored 74: MIT, inside a repository moving new contributions to Apache-2.0, both OSI licences (28). Local software with no network calls. The README says where the graph is stored, and the code shows the whole file is rewritten on each change (20). A legacy memory.json is migrated to memory.jsonl automatically, but there's no deprecation policy or notice with dates (6). No telemetry in the code (20). The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency): - 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms. - 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors). - 0 to 20, a deprecation policy or notices with dates. - 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted). The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two. Provenance checks not met in full (half of this category, computed from checked facts): - Domain age: modelcontextprotocol.io, registered 2024-11-18 (1 year) (3 of 15) - Status page: not found (0 of 10) ## What we couldn't check What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it. - unchecked: whether the TypeScript CI workflow passes on the current head of main - unchecked: open memory issues and reply times, since GitHub's robots rules blocked issue search - When the eleven September memory fixes will be released ## Weaknesses - No pagination or limits. `read_graph` returns everything and `search_nodes` every match - The published version can lose one of two writes made in the same turn. The fix is merged but unreleased - Search is case-insensitive substring matching, with no ranking or semantics - The default file location is inside the package directory, where a reinstall or cache clean can remove it - No read-only mode, and stored text returns verbatim in later sessions ## What costs an agent a turn today The notes we give agents before they call it. Each one is a workaround an agent shouldn't need. - Set `MEMORY_FILE_PATH` to an absolute path. The default sits inside the npx cache - Use `search_nodes` or `open_nodes` instead of `read_graph` once the graph has more than a few hundred entities - Make memory writes one at a time. Parallel calls in one turn can overwrite each other in 2026.8.31 - Create both entities before `create_relations`. A missing endpoint fails the whole batch - Keep observations short and factual. They come back verbatim in every later read ## What the review panel asked for - release the September fixes - say when to use search_nodes - read-only mode - provenance on stored facts ## When it's done Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.