{
  "fixes": {
    "slug": "graphicle",
    "name": "Graphicle",
    "listing": "https://www.anchorterminal.com/tools/graphicle",
    "markdown": "# Fix list: Graphicle\n\nFrom Anchor Terminal's listing at https://www.anchorterminal.com/tools/graphicle, the October 2026 research run, assessed 15 October 2026. Grade B, 69.8 out of 100.\n\nThis 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.\n\nFor a coding agent working on Graphicle: 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.\n\n## 1. Reliability, 63 out of 100, up to 7.4 more on the total\n\nWhy it scored 63: Read as local software. Installs from an official Homebrew tap, an install script that checks each archive's SHA-256, or release archives, with macOS and Linux (glibc) on arm64 and x86_64 stated. No language package registry (20). A public CI workflow formats, lints and tests each of the three crates on macOS and Linux, and every test job passed on both pushes to main. The cargo-deny job for licences, advisories and sources failed on both, and its log was not readable without a GitHub login (18). No issues filed in the project's first day. The only change since release is an outside fix for documents whose text carries a NUL byte, so the record is too short to judge (15). A dated changelog with semver crate versions, one release so far, so no breaking change to call out yet (10). Version 0.1.0, not declared stable (0).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):\n\nHosted APIs, MCP servers, models and platforms.\n\n- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).\n- 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.\n- 15, rate limits documented with numbers.\n- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.\n- 10, an SLA published for any paid tier.\n- 10, the surface agents use is generally available, not beta or preview.\n\nLocal packages, SDKs, frameworks and stdio MCP servers.\n\n- 20, installs from an official package with supported runtimes stated.\n- 25, a public CI and test suite, passing on the default branch.\n- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).\n- 15, semver discipline and breaking changes called out in a changelog.\n- 15, version 1.0 or later, or declared stable.\n\nProtocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.\n\n## 2. Security \u0026 auth, 69 out of 100, up to 5.4 more on the total\n\nWhy it scored 69: Read with the API checklist for a server the owner runs, as for Cognee and Graphiti. Over HTTP each person or agent gets its own key with a role, optional expiry and optional project list, stored as a hash, revocable, and sent in a header with the docs saying never in the URL. The stdio server runs with the owner key, which can read every project, and its graph tools take `repo` to reach other projects (24 of 30). The MCP server has no delete or command tool, agent writes are drafts that a person approves, and `--tools` and per-key project revocation narrow access. No approval step is needed because nothing destructive is exposed (18 of 20). Recall returns documents, connector content and captured sessions, which can carry injected text. Server instructions say memory is evidence, not tool permission, drafts need verbatim quotes, and capture redacts credentials and drops `\u003cprivate\u003e` spans. No general prompt-injection guidance found (9 of 15). Workspace usage counters, `workflow_status` and hash-chained deletion receipts, while the team audit trail belongs to the commercial edition (6 of 15). SECURITY.md with private reporting through GitHub advisories, a valid security.txt at karlex.ai and a cargo-deny advisory check in CI that currently fails. No bug bounty or certification (10 of 20).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-security):\n\n- 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.\n- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.\n- 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.\n- 0 to 15, audit logs or per-call visibility for the operator.\n- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.\n\nModels 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.\n\n## 3. Payments \u0026 pricing, 60 out of 100, up to 5 more on the total\n\nWhy it scored 60: No payment protocol (0). Read with the self-hosted rule. Free under Apache-2.0 with nothing to buy (20). No card or account (20). An agent can install it and run `graphicle setup --yes` with no signup, as the README suggests (20). The commercial team edition is in private preview with no prices, so there is no paid option to score.\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):\n\nThe published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).\n\n- 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.\n- 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.\n- 20, a free tier or trial that doesn't need a card.\n- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).\n\nPayment 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.\n\nOpen-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.\n\n## 4. Maintenance \u0026 community, 48 out of 100, up to 4.6 more on the total\n\nWhy it scored 48: v0.1.0 on 14 October 2026 (30). One release in the last 90 days (0). An outside pull request was merged about three and a half hours after it opened, with a maintainer reply, and Discussions and a Discord are linked. One day of history (12 of 25). Not in the official MCP registry, which returned nothing for graphicle on 15 October, and the SDKs are not published (0). Pinned GitHub Actions, dependency-graph updates and passing tests, but the licence and advisory job fails on main (6 of 10).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):\n\n- 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.\n- 20, at least three releases or dated changelog entries in the last 90 days.\n- 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.\n- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).\n- 10, package health, current dependencies and CI.\n\nModels are read for deprecation notice periods and model churn rather than release counts.\n\n## 5. Schema \u0026 documentation, 83 out of 100, up to 2.8 more on the total\n\nWhy it scored 83: Every MCP tool has a JSON Schema input with `additionalProperties: false` and required fields marked, read from the tool list fixture in the repository (25). No llms.txt. The docs are Markdown files in the repository that an agent can read raw (5 of 10). Server instructions say which tool to use for structure and which for what the project says, `recall_project` says when to leave the reranker off, and memory tools explain draft review, while graph tools such as `query_graph` carry one-line descriptions (14 of 20). Enums on `mode`, `pack_order`, memory `kind` and `state`, length and item limits, minimum values and closed nested objects (14 of 15). Examples throughout the docs. The OpenAPI file documents 400, 403, 409 and 429 for the HTTP API and limit errors name the setting to change, but MCP error results are not catalogued (10 of 15). A dated changelog, semver and `/v1` paths (15).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):\n\nAPIs and MCP servers.\n\n- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).\n- 10, llms.txt or Markdown docs served for agents.\n- 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.\n- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.\n- 0 to 15, examples and documented error responses.\n- 15, versioning and a public changelog.\n\nModels 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.\n\n## 6. Transparency \u0026 trust, 68 out of 100, up to 2.8 more on the total\n\nMade of editorial 71, provenance 65.\n\nWhy it scored 68: Apache-2.0 with NOTICE files that credit Graphify and every adapted project (30). The README privacy section and the FAQ list every case in which the program uses the network, say there is no telemetry or account, and keep data in one SQLite store with deletion receipts. A grep of the source for outbound hosts found only model providers, connectors and GitHub, which agrees. The karlex.ai privacy policy of 14 March 2025 covers the website, not the software (20 of 30). SECURITY.md lists supported versions. No deprecation policy (3 of 20). No telemetry to opt out of, and the network cases are disclosed, with `freshness off` and `--code-only` to limit them (18 of 20).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):\n\n- 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.\n- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).\n- 0 to 20, a deprecation policy or notices with dates.\n- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).\n\nThe other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.\n\nProvenance checks not met in full (half of this category, computed from checked facts):\n\n- Domain age: karlex.ai, no registry record we could read (0 of 15)\n- Terms of service: read, states 6 of the 7 things a reader expects, and has 2 clauses that cost points (5.1 of 10)\n- Status page: not found (0 of 10)\n\n## 7. Agent ergonomics, 86 out of 100, up to 2.3 more on the total\n\nWhy it scored 86: 25 tools with memory proposals on, which `graphicle setup` turns on by default, about 24,000 characters of definitions (15). `--tools graph` or `--tools memory` serves one half, and leaving out `--allow-learning` leaves only read tools plus `graph_link` (7 added back). `recall_project` takes `max_tokens` or `max_chars`, `include` prefixes and `skip_shown`, `query_graph` takes `token_budget` and `depth`, and `list_projects` pages with `offset` and `limit` (20). Over-limit projects are refused with a message naming the limit and its setting, 409 marks a stale revision and 429 carries Retry-After. MCP error shapes are not documented (12 of 20). Every tool carries readOnlyHint, idempotentHint and destructiveHint false, there are no destructive tools, and `propose_memory` requires an `idempotency_key` that replays the same proposal (20). Only `question` is required on recall and graph queries. Python and TypeScript SDKs exist but install only from source (8 of 15).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):\n\n- 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).\n- 20, pagination, filtering and output-size controls.\n- 20, actionable, documented error responses, codes and messages an agent can recover from.\n- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.\n- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.\n\nModels 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.\n\n## What we couldn't check\n\nWhat 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.\n\n- Judgement call. Listed in agent-memory, not code. Its main job is recall of a project's sources and reviewed memory, beside Graphiti and Cognee, and the code graph feeds that recall. The code category covers repository platforms and library documentation.\n- The lead was right. The changelog calls 0.1.0 of 14 October 2026 the first public release, and the repository was created that day. The project was one day old when graded.\n- The lead's \"SDKs\" are Python and TypeScript source packages in `sdk/`, not published to PyPI or npm.\n- unchecked: why the cargo-deny job (licences, advisories, sources) fails. GitHub job logs need a login.\n- We did not install or run the program. Behaviour comes from the source, docs, tool list fixture and changelog. Speed and memory benchmark figures are the vendor's.\n- KarLex AI's about page says an earlier project, ASIOS, was announced in May 2025 and discontinued.\n\n## Weaknesses\n\n- One release, v0.1.0 on 14 October 2026, from a repository created that day\n- The cargo-deny licence and advisory job failed on both pushes to main, though the test jobs passed on macOS and Linux\n- The stdio server runs with the owner key, so its graph tools can read any project in the store unless an HTTP key is scoped\n- SDKs are source-only, no llms.txt, not in the official MCP registry, and no native Windows build\n- The team audit trail is in the commercial edition. Core shows usage counters and deletion receipts only\n\n## What costs an agent a turn today\n\nThe notes we give agents before they call it. Each one is a workaround an agent shouldn't need.\n\n- Call `recall_project` with a `max_tokens` budget, then `read_project_source` with the cited id and byte offset to continue a cut excerpt\n- Use `query_graph`, `get_neighbors` and `shortest_path` for structure questions, and pass `repo` to reach another synced project\n- Treat proposed memories as drafts. They are not served until a person approves them in `graphicle open` under Memory\n- Ask the user before running `graphicle setup`, which edits agent settings, AGENTS.md and Git hooks, and use `graphicle disconnect` to undo it\n- Over HTTP send the key as `Authorization: Bearer`, never in the URL, and on 429 wait for the Retry-After reset\n\n## When it's done\n\nSend 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.\n",
    "grade": "B",
    "score": 69.8,
    "assessed": "2026-10-15",
    "run": "October 2026 research run",
    "categories": [
      {
        "key": "reliability",
        "name": "Reliability",
        "score": 63,
        "maxGain": 7.4,
        "reason": "Read as local software. Installs from an official Homebrew tap, an install script that checks each archive's SHA-256, or release archives, with macOS and Linux (glibc) on arm64 and x86_64 stated. No language package registry (20). A public CI workflow formats, lints and tests each of the three crates on macOS and Linux, and every test job passed on both pushes to main. The cargo-deny job for licences, advisories and sources failed on both, and its log was not readable without a GitHub login (18). No issues filed in the project's first day. The only change since release is an outside fix for documents whose text carries a NUL byte, so the record is too short to judge (15). A dated changelog with semver crate versions, one release so far, so no breaking change to call out yet (10). Version 0.1.0, not declared stable (0).",
        "checklist": [
          "Hosted APIs, MCP servers, models and platforms.",
          "- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).\n- 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.\n- 15, rate limits documented with numbers.\n- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.\n- 10, an SLA published for any paid tier.\n- 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.\n- 25, a public CI and test suite, passing on the default branch.\n- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).\n- 15, semver discipline and breaking changes called out in a changelog.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-reliability"
      },
      {
        "key": "security",
        "name": "Security \u0026 auth",
        "score": 69,
        "maxGain": 5.4,
        "reason": "Read with the API checklist for a server the owner runs, as for Cognee and Graphiti. Over HTTP each person or agent gets its own key with a role, optional expiry and optional project list, stored as a hash, revocable, and sent in a header with the docs saying never in the URL. The stdio server runs with the owner key, which can read every project, and its graph tools take `repo` to reach other projects (24 of 30). The MCP server has no delete or command tool, agent writes are drafts that a person approves, and `--tools` and per-key project revocation narrow access. No approval step is needed because nothing destructive is exposed (18 of 20). Recall returns documents, connector content and captured sessions, which can carry injected text. Server instructions say memory is evidence, not tool permission, drafts need verbatim quotes, and capture redacts credentials and drops `\u003cprivate\u003e` spans. No general prompt-injection guidance found (9 of 15). Workspace usage counters, `workflow_status` and hash-chained deletion receipts, while the team audit trail belongs to the commercial edition (6 of 15). SECURITY.md with private reporting through GitHub advisories, a valid security.txt at karlex.ai and a cargo-deny advisory check in CI that currently fails. No bug bounty or certification (10 of 20).",
        "checklist": [
          "- 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.\n- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.\n- 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.\n- 0 to 15, audit logs or per-call visibility for the operator.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-security"
      },
      {
        "key": "payments",
        "name": "Payments \u0026 pricing",
        "score": 60,
        "maxGain": 5,
        "reason": "No payment protocol (0). Read with the self-hosted rule. Free under Apache-2.0 with nothing to buy (20). No card or account (20). An agent can install it and run `graphicle setup --yes` with no signup, as the README suggests (20). The commercial team edition is in private preview with no prices, so there is no paid option to score.",
        "checklist": [
          "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.\n- 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.\n- 20, a free tier or trial that doesn't need a card.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-payments"
      },
      {
        "key": "maintenance",
        "name": "Maintenance \u0026 community",
        "score": 48,
        "maxGain": 4.6,
        "reason": "v0.1.0 on 14 October 2026 (30). One release in the last 90 days (0). An outside pull request was merged about three and a half hours after it opened, with a maintainer reply, and Discussions and a Discord are linked. One day of history (12 of 25). Not in the official MCP registry, which returned nothing for graphicle on 15 October, and the SDKs are not published (0). Pinned GitHub Actions, dependency-graph updates and passing tests, but the licence and advisory job fails on main (6 of 10).",
        "checklist": [
          "- 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.\n- 20, at least three releases or dated changelog entries in the last 90 days.\n- 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.\n- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).\n- 10, package health, current dependencies and CI.",
          "Models are read for deprecation notice periods and model churn rather than release counts."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-maintenance"
      },
      {
        "key": "schema",
        "name": "Schema \u0026 documentation",
        "score": 83,
        "maxGain": 2.8,
        "reason": "Every MCP tool has a JSON Schema input with `additionalProperties: false` and required fields marked, read from the tool list fixture in the repository (25). No llms.txt. The docs are Markdown files in the repository that an agent can read raw (5 of 10). Server instructions say which tool to use for structure and which for what the project says, `recall_project` says when to leave the reranker off, and memory tools explain draft review, while graph tools such as `query_graph` carry one-line descriptions (14 of 20). Enums on `mode`, `pack_order`, memory `kind` and `state`, length and item limits, minimum values and closed nested objects (14 of 15). Examples throughout the docs. The OpenAPI file documents 400, 403, 409 and 429 for the HTTP API and limit errors name the setting to change, but MCP error results are not catalogued (10 of 15). A dated changelog, semver and `/v1` paths (15).",
        "checklist": [
          "APIs and MCP servers.",
          "- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).\n- 10, llms.txt or Markdown docs served for agents.\n- 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.\n- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.\n- 0 to 15, examples and documented error responses.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-schema"
      },
      {
        "key": "transparency",
        "name": "Transparency \u0026 trust",
        "score": 68,
        "maxGain": 2.8,
        "reason": "Apache-2.0 with NOTICE files that credit Graphify and every adapted project (30). The README privacy section and the FAQ list every case in which the program uses the network, say there is no telemetry or account, and keep data in one SQLite store with deletion receipts. A grep of the source for outbound hosts found only model providers, connectors and GitHub, which agrees. The karlex.ai privacy policy of 14 March 2025 covers the website, not the software (20 of 30). SECURITY.md lists supported versions. No deprecation policy (3 of 20). No telemetry to opt out of, and the network cases are disclosed, with `freshness off` and `--code-only` to limit them (18 of 20).",
        "blend": "editorial 71, provenance 65",
        "checklist": [
          "- 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.\n- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).\n- 0 to 20, a deprecation policy or notices with dates.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-transparency"
      },
      {
        "key": "ergonomics",
        "name": "Agent ergonomics",
        "score": 86,
        "maxGain": 2.3,
        "reason": "25 tools with memory proposals on, which `graphicle setup` turns on by default, about 24,000 characters of definitions (15). `--tools graph` or `--tools memory` serves one half, and leaving out `--allow-learning` leaves only read tools plus `graph_link` (7 added back). `recall_project` takes `max_tokens` or `max_chars`, `include` prefixes and `skip_shown`, `query_graph` takes `token_budget` and `depth`, and `list_projects` pages with `offset` and `limit` (20). Over-limit projects are refused with a message naming the limit and its setting, 409 marks a stale revision and 429 carries Retry-After. MCP error shapes are not documented (12 of 20). Every tool carries readOnlyHint, idempotentHint and destructiveHint false, there are no destructive tools, and `propose_memory` requires an `idempotency_key` that replays the same proposal (20). Only `question` is required on recall and graph queries. Python and TypeScript SDKs exist but install only from source (8 of 15).",
        "checklist": [
          "- 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).\n- 20, pagination, filtering and output-size controls.\n- 20, actionable, documented error responses, codes and messages an agent can recover from.\n- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-ergonomics"
      }
    ],
    "provenance": [
      {
        "label": "Domain age",
        "value": "karlex.ai, no registry record we could read",
        "points": 0,
        "max": 15
      },
      {
        "label": "Terms of service",
        "value": "read, states 6 of the 7 things a reader expects, and has 2 clauses that cost points",
        "points": 5.1,
        "max": 10
      },
      {
        "label": "Status page",
        "value": "not found",
        "points": 0,
        "max": 10
      }
    ],
    "unchecked": [
      "Judgement call. Listed in agent-memory, not code. Its main job is recall of a project's sources and reviewed memory, beside Graphiti and Cognee, and the code graph feeds that recall. The code category covers repository platforms and library documentation.",
      "The lead was right. The changelog calls 0.1.0 of 14 October 2026 the first public release, and the repository was created that day. The project was one day old when graded.",
      "The lead's \"SDKs\" are Python and TypeScript source packages in `sdk/`, not published to PyPI or npm.",
      "unchecked: why the cargo-deny job (licences, advisories, sources) fails. GitHub job logs need a login.",
      "We did not install or run the program. Behaviour comes from the source, docs, tool list fixture and changelog. Speed and memory benchmark figures are the vendor's.",
      "KarLex AI's about page says an earlier project, ASIOS, was announced in May 2025 and discontinued."
    ],
    "weaknesses": [
      "One release, v0.1.0 on 14 October 2026, from a repository created that day",
      "The cargo-deny licence and advisory job failed on both pushes to main, though the test jobs passed on macOS and Linux",
      "The stdio server runs with the owner key, so its graph tools can read any project in the store unless an HTTP key is scoped",
      "SDKs are source-only, no llms.txt, not in the official MCP registry, and no native Windows build",
      "The team audit trail is in the commercial edition. Core shows usage counters and deletion receipts only"
    ],
    "agentNotes": [
      "Call `recall_project` with a `max_tokens` budget, then `read_project_source` with the cited id and byte offset to continue a cut excerpt",
      "Use `query_graph`, `get_neighbors` and `shortest_path` for structure questions, and pass `repo` to reach another synced project",
      "Treat proposed memories as drafts. They are not served until a person approves them in `graphicle open` under Memory",
      "Ask the user before running `graphicle setup`, which edits agent settings, AGENTS.md and Git hooks, and use `graphicle disconnect` to undo it",
      "Over HTTP send the key as `Authorization: Bearer`, never in the URL, and on 429 wait for the Retry-After reset"
    ],
    "recheck": "https://www.anchorterminal.com/builders/#disputes"
  },
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "attributionRequired": true,
    "credit": "CC BY 4.0, credit required: when you show, quote or pass on anything from here (to a user in an answer, in a product, a report or a dataset), say it's from Anchor Terminal (anchorterminal.com) and link the page or API address you took it from. Reviews also credit their author.",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-10-15",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "0.4",
    "openapi": "https://www.anchorterminal.com/openapi.json",
    "preview": false,
    "run": "2026-10-01",
    "runLabel": "October 2026 research run",
    "terms": "https://www.anchorterminal.com/terms/#data"
  }
}