{
  "fixes": {
    "slug": "turnkey-agentic-wallets",
    "name": "Turnkey Agentic Wallets",
    "listing": "https://www.anchorterminal.com/tools/turnkey-agentic-wallets",
    "markdown": "# Fix list: Turnkey Agentic Wallets\n\nFrom Anchor Terminal's listing at https://www.anchorterminal.com/tools/turnkey-agentic-wallets, the October 2026 research run, assessed 8 October 2026. Grade BB, 76 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 Turnkey Agentic Wallets: 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. Payments \u0026 pricing, 58 out of 100, up to 5.3 more on the total\n\nWhy it scored 58: Payment platforms and wallets take the highest step that applies on the 40-point protocol line. 40 for x402, MPP or L402 on all their own endpoints, 30 on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, 0 for only a protocol of their own. Turnkey documents x402 and MPP clients signing through `@turnkey/viem`, and its own API is not paid over either, so the buyer step (15 of 40). Prices are public per signature, $0.10 on Pay as You Go and $0.05 on Pro with a $99 monthly minimum, while transaction management and swap fees need a sales contact (18 of 20). 25 free signatures a month, with a card needed only beyond that (20). A person creates the organisation in the dashboard with an email address and a passkey. After that, users, keys, wallets and policies are created through the API (5 of 20).\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## 2. Reliability, 75 out of 100, up to 5 more on the total\n\nWhy it scored 75: Status page at turnkey-status.com on Atlassian Statuspage with 23 components and 90-day uptime bars (20). Between 10 July and 8 October 2026 it lists seven incidents. Signing endpoints, `sign_raw_payload` and `sign_transaction` among them, returned elevated errors for 5 hours 9 minutes on 27 August, marked minor. SMS authentication was down for about 12 hours on 6 and 7 August, marked critical and outside the path an agent's API key uses. The rest were webhook delays and broadcast failures traced to upstream RPC providers (12 of 30). Rate limits are published as 1 request a second on Free and Pay as You Go, 3 on Pro and 60 on Enterprise, with 10 a second for each sub-organisation and for broadcast and balance calls (15). The docs describe 429 handling with exponential backoff and jitter and say submissions are idempotent by body fingerprint, but the API returns no `Retry-After` header (13 of 15). SLAs are listed on the Enterprise plan with no published terms (5 of 10). The API is generally available and the agent skills are at 1.0.0 (10).\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## 3. Agent ergonomics, 77 out of 100, up to 3.7 more on the total\n\nWhy it scored 77: There is no MCP server for wallet actions; the MCP server at docs.turnkey.com/mcp searches the docs only. An agent calls the API through an SDK, the CLI or the skills, and responses can't be trimmed by field (15 of 25). List queries take cursor pagination with a limit of 1 to 100 and filters by activity status and type (17 of 20). Errors carry a category code and a readable message with documented fixes, wrapped in a gRPC status the docs say not to rely on (16 of 20). Submissions are idempotent by a hash of the POST body, and the docs explain that changing `timestampMs` creates a new activity. No MCP annotations apply (18 of 20). Server SDKs exist for TypeScript, Go, Ruby, Rust and Python, with wallet and policy helpers in the first three, but every request needs a P-256 signature in an `X-Stamp` header and submissions need an activity envelope (11 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## 4. Schema \u0026 documentation, 79 out of 100, up to 3.4 more on the total\n\nWhy it scored 79: A public Swagger 2.0 document at docs.turnkey.com/public_api.swagger.json with 186 paths, 71 queries and 114 submissions (25). llms.txt, llms-full.txt and a Markdown twin of every docs page (10). Operation descriptions in the spec are one line each, such as \"Sign a transaction.\", while the guides and the nine agent skills say when to use each call and when not to (13 of 20). The spec has 175 enums and required fields on every request, but 64-bit numbers travel as strings and policy `condition` and `consensus` are free-form strings in Turnkey's own expression language (11 of 15). The errors page lists 32 `turnkeyErrorCode` values and a table of messages with HTTP codes and fixes, though the spec documents only the 200 response for each path (12 of 15). Paths sit under `/public/v1` and activity types carry version suffixes, but no changelog for the API was found in the docs. Changes appear only in the SDK changelogs on GitHub (8 of 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## 5. Security \u0026 auth, 85 out of 100, up to 2.6 more on the total\n\nWhy it scored 85: Each user holds P-256 API key pairs whose private half is generated by the customer and never sent to Turnkey. Keys can be created with an expiry, a user has up to ten of each kind, deleting the user revokes access at once, and no secret travels in a URL. Scope comes from policies, not from the key itself (28 of 30). Policies deny by default, DENY beats ALLOW, a non-root user starts with no permissions, read access needs no policy, and consensus expressions can require a named person to approve. Root users bypass the policy engine, and no daily or rolling spending limit was found in the policy language, only conditions on a single transaction (17 of 20). The skills page warns that models can misread instructions and that policies are the technical control, and the terms say the same. No guidance on untrusted content in balances or token data was found (9 of 15). Every action is an activity recording who approved it, readable by query and delivered by signed webhooks (13 of 15). SOC 2 Type II per the docs, a bug bounty of up to $50,000 with safe harbour, a security.txt with a PGP key and no Expires field, a published whitepaper and the enclave operating system QuorumOS open source (18 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## 6. Transparency \u0026 trust, 72 out of 100, up to 2.5 more on the total\n\nMade of editorial 50, provenance 94.\n\nWhy it scored 72: A closed service under published terms of service, updated 6 August 2026, with the SDKs, CLI and agent skills under Apache-2.0 and QuorumOS open source (18 of 30). The privacy policy, updated 29 June 2026, names Turnkey Global, Inc., says Turnkey is a processor for its business customers' users and lists categories of service providers. It gives no retention periods, and the data processing terms were not read because the trust centre is drawn by script (17 of 30). Deprecated activity types are marked in the docs with migration guides, with no dates or notice period found (7 of 20). The docs say subprocessors and their locations are in the trust centre, which we couldn't read. The status page names AWS SES and SNS in us-east-1, Cloudflare and Alchemy as dependencies (8 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- Terms of service: read, states 7 of the 7 things a reader expects, and has 3 clauses that cost points (4 of 10)\n\n## 7. Maintenance \u0026 community, 83 out of 100, up to 1.5 more on the total\n\nWhy it scored 83: The SDK repository tagged v2026.10.0 on 8 October 2026 and `@turnkey/sdk-server` 8.6.0 on 22 September (30). Twelve dated release tags between 6 August and 8 October (20). The service is closed, with dashboard chat, email and a community Slack for support and pull requests merged on the day of the last tag. We didn't test a support channel and no API changelog was found (10 of 15). Current official SDKs in TypeScript, Go, Ruby, Rust and Python (15). Public CI workflows, a dependency patch merged on 8 October and Renovate on the skills repository. The CLI's last tag is May 2025 (8 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## 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- unchecked: the trust centre at trust.turnkey.com is drawn by script, so the subprocessor list, data locations, the SOC 2 report details and any data processing addendum were not read\n- unchecked: GitHub star counts, because the GitHub API refused our requests for the rate limit\n- unchecked: open issues and reply times on tkhq/sdk, and whether a support channel answers\n- unchecked: whether signing up needs a card. The pricing FAQ says a card is needed only past 25 signatures\n- unchecked: the tools the docs MCP server lists. We did not connect to it\n- No daily, rolling or cumulative spending limit was found in the policy language. Turnkey's marketing for a customer mentions tamper-proof spend limits, and how those are built was not established\n- The rate limits page names a Free plan that the pricing page does not list, and the pricing FAQ says Pro can come down to $0.01 a signature where the table says $0.05\n- The terms forbid using the services for availability or performance monitoring and for benchmarking, which matters before our probes run\n- The status page's robots.txt disallows `/api/`. One request reached `/api/v2/incidents.json` before we read robots.txt. The file was deleted unread and the incident record here comes from the `/history` pages\n- llms.txt and every Markdown docs page open with a block headed Agent Instructions addressed to AI models. We recorded it as a fact and did not act on it\n- Agentic Wallets is a documented use of Turnkey's general wallet API and not a separate endpoint, plan or price\n\n## Weaknesses\n\n- Policy conditions cap value per transaction. No daily, rolling or cumulative spending limit was found in the policy language\n- Free, Pay as You Go and Pro organisations are limited to five policies and a default of 1 to 3 requests a second\n- Root users bypass the policy engine entirely, and the first user of a new organisation is root\n- Signing endpoints returned elevated errors from 00:44 to 05:53 EDT on 27 August 2026\n- 429 responses carry no `Retry-After` header, and the terms forbid benchmarking and availability monitoring\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- Use a non-root user's key. Refuse root credentials for unattended work, because root bypasses every policy\n- Scope every signing ALLOW policy by `wallet.id` or `wallet_account.address`, or it covers every key the user can reach\n- Check the activity status on each response. `ACTIVITY_STATUS_CONSENSUS_NEEDED` means a person must approve within 24 hours\n- Retry with the identical body to stay idempotent. Changing `timestampMs` creates a new activity\n- Back off with jitter on 429 and poll `get_activity` sparingly, since polling counts against the same 1 to 3 requests a second\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": "BB",
    "score": 76,
    "assessed": "2026-10-08",
    "run": "October 2026 research run",
    "categories": [
      {
        "key": "payments",
        "name": "Payments \u0026 pricing",
        "score": 58,
        "maxGain": 5.3,
        "reason": "Payment platforms and wallets take the highest step that applies on the 40-point protocol line. 40 for x402, MPP or L402 on all their own endpoints, 30 on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, 0 for only a protocol of their own. Turnkey documents x402 and MPP clients signing through `@turnkey/viem`, and its own API is not paid over either, so the buyer step (15 of 40). Prices are public per signature, $0.10 on Pay as You Go and $0.05 on Pro with a $99 monthly minimum, while transaction management and swap fees need a sales contact (18 of 20). 25 free signatures a month, with a card needed only beyond that (20). A person creates the organisation in the dashboard with an email address and a passkey. After that, users, keys, wallets and policies are created through the API (5 of 20).",
        "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": "reliability",
        "name": "Reliability",
        "score": 75,
        "maxGain": 5,
        "reason": "Status page at turnkey-status.com on Atlassian Statuspage with 23 components and 90-day uptime bars (20). Between 10 July and 8 October 2026 it lists seven incidents. Signing endpoints, `sign_raw_payload` and `sign_transaction` among them, returned elevated errors for 5 hours 9 minutes on 27 August, marked minor. SMS authentication was down for about 12 hours on 6 and 7 August, marked critical and outside the path an agent's API key uses. The rest were webhook delays and broadcast failures traced to upstream RPC providers (12 of 30). Rate limits are published as 1 request a second on Free and Pay as You Go, 3 on Pro and 60 on Enterprise, with 10 a second for each sub-organisation and for broadcast and balance calls (15). The docs describe 429 handling with exponential backoff and jitter and say submissions are idempotent by body fingerprint, but the API returns no `Retry-After` header (13 of 15). SLAs are listed on the Enterprise plan with no published terms (5 of 10). The API is generally available and the agent skills are at 1.0.0 (10).",
        "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": "ergonomics",
        "name": "Agent ergonomics",
        "score": 77,
        "maxGain": 3.7,
        "reason": "There is no MCP server for wallet actions; the MCP server at docs.turnkey.com/mcp searches the docs only. An agent calls the API through an SDK, the CLI or the skills, and responses can't be trimmed by field (15 of 25). List queries take cursor pagination with a limit of 1 to 100 and filters by activity status and type (17 of 20). Errors carry a category code and a readable message with documented fixes, wrapped in a gRPC status the docs say not to rely on (16 of 20). Submissions are idempotent by a hash of the POST body, and the docs explain that changing `timestampMs` creates a new activity. No MCP annotations apply (18 of 20). Server SDKs exist for TypeScript, Go, Ruby, Rust and Python, with wallet and policy helpers in the first three, but every request needs a P-256 signature in an `X-Stamp` header and submissions need an activity envelope (11 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"
      },
      {
        "key": "schema",
        "name": "Schema \u0026 documentation",
        "score": 79,
        "maxGain": 3.4,
        "reason": "A public Swagger 2.0 document at docs.turnkey.com/public_api.swagger.json with 186 paths, 71 queries and 114 submissions (25). llms.txt, llms-full.txt and a Markdown twin of every docs page (10). Operation descriptions in the spec are one line each, such as \"Sign a transaction.\", while the guides and the nine agent skills say when to use each call and when not to (13 of 20). The spec has 175 enums and required fields on every request, but 64-bit numbers travel as strings and policy `condition` and `consensus` are free-form strings in Turnkey's own expression language (11 of 15). The errors page lists 32 `turnkeyErrorCode` values and a table of messages with HTTP codes and fixes, though the spec documents only the 200 response for each path (12 of 15). Paths sit under `/public/v1` and activity types carry version suffixes, but no changelog for the API was found in the docs. Changes appear only in the SDK changelogs on GitHub (8 of 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": "security",
        "name": "Security \u0026 auth",
        "score": 85,
        "maxGain": 2.6,
        "reason": "Each user holds P-256 API key pairs whose private half is generated by the customer and never sent to Turnkey. Keys can be created with an expiry, a user has up to ten of each kind, deleting the user revokes access at once, and no secret travels in a URL. Scope comes from policies, not from the key itself (28 of 30). Policies deny by default, DENY beats ALLOW, a non-root user starts with no permissions, read access needs no policy, and consensus expressions can require a named person to approve. Root users bypass the policy engine, and no daily or rolling spending limit was found in the policy language, only conditions on a single transaction (17 of 20). The skills page warns that models can misread instructions and that policies are the technical control, and the terms say the same. No guidance on untrusted content in balances or token data was found (9 of 15). Every action is an activity recording who approved it, readable by query and delivered by signed webhooks (13 of 15). SOC 2 Type II per the docs, a bug bounty of up to $50,000 with safe harbour, a security.txt with a PGP key and no Expires field, a published whitepaper and the enclave operating system QuorumOS open source (18 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": "transparency",
        "name": "Transparency \u0026 trust",
        "score": 72,
        "maxGain": 2.5,
        "reason": "A closed service under published terms of service, updated 6 August 2026, with the SDKs, CLI and agent skills under Apache-2.0 and QuorumOS open source (18 of 30). The privacy policy, updated 29 June 2026, names Turnkey Global, Inc., says Turnkey is a processor for its business customers' users and lists categories of service providers. It gives no retention periods, and the data processing terms were not read because the trust centre is drawn by script (17 of 30). Deprecated activity types are marked in the docs with migration guides, with no dates or notice period found (7 of 20). The docs say subprocessors and their locations are in the trust centre, which we couldn't read. The status page names AWS SES and SNS in us-east-1, Cloudflare and Alchemy as dependencies (8 of 20).",
        "blend": "editorial 50, provenance 94",
        "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": "maintenance",
        "name": "Maintenance \u0026 community",
        "score": 83,
        "maxGain": 1.5,
        "reason": "The SDK repository tagged v2026.10.0 on 8 October 2026 and `@turnkey/sdk-server` 8.6.0 on 22 September (30). Twelve dated release tags between 6 August and 8 October (20). The service is closed, with dashboard chat, email and a community Slack for support and pull requests merged on the day of the last tag. We didn't test a support channel and no API changelog was found (10 of 15). Current official SDKs in TypeScript, Go, Ruby, Rust and Python (15). Public CI workflows, a dependency patch merged on 8 October and Renovate on the skills repository. The CLI's last tag is May 2025 (8 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"
      }
    ],
    "provenance": [
      {
        "label": "Terms of service",
        "value": "read, states 7 of the 7 things a reader expects, and has 3 clauses that cost points",
        "points": 4,
        "max": 10
      }
    ],
    "unchecked": [
      "unchecked: the trust centre at trust.turnkey.com is drawn by script, so the subprocessor list, data locations, the SOC 2 report details and any data processing addendum were not read",
      "unchecked: GitHub star counts, because the GitHub API refused our requests for the rate limit",
      "unchecked: open issues and reply times on tkhq/sdk, and whether a support channel answers",
      "unchecked: whether signing up needs a card. The pricing FAQ says a card is needed only past 25 signatures",
      "unchecked: the tools the docs MCP server lists. We did not connect to it",
      "No daily, rolling or cumulative spending limit was found in the policy language. Turnkey's marketing for a customer mentions tamper-proof spend limits, and how those are built was not established",
      "The rate limits page names a Free plan that the pricing page does not list, and the pricing FAQ says Pro can come down to $0.01 a signature where the table says $0.05",
      "The terms forbid using the services for availability or performance monitoring and for benchmarking, which matters before our probes run",
      "The status page's robots.txt disallows `/api/`. One request reached `/api/v2/incidents.json` before we read robots.txt. The file was deleted unread and the incident record here comes from the `/history` pages",
      "llms.txt and every Markdown docs page open with a block headed Agent Instructions addressed to AI models. We recorded it as a fact and did not act on it",
      "Agentic Wallets is a documented use of Turnkey's general wallet API and not a separate endpoint, plan or price"
    ],
    "weaknesses": [
      "Policy conditions cap value per transaction. No daily, rolling or cumulative spending limit was found in the policy language",
      "Free, Pay as You Go and Pro organisations are limited to five policies and a default of 1 to 3 requests a second",
      "Root users bypass the policy engine entirely, and the first user of a new organisation is root",
      "Signing endpoints returned elevated errors from 00:44 to 05:53 EDT on 27 August 2026",
      "429 responses carry no `Retry-After` header, and the terms forbid benchmarking and availability monitoring"
    ],
    "agentNotes": [
      "Use a non-root user's key. Refuse root credentials for unattended work, because root bypasses every policy",
      "Scope every signing ALLOW policy by `wallet.id` or `wallet_account.address`, or it covers every key the user can reach",
      "Check the activity status on each response. `ACTIVITY_STATUS_CONSENSUS_NEEDED` means a person must approve within 24 hours",
      "Retry with the identical body to stay idempotent. Changing `timestampMs` creates a new activity",
      "Back off with jitter on 429 and poll `get_activity` sparingly, since polling counts against the same 1 to 3 requests a second"
    ],
    "recheck": "https://www.anchorterminal.com/builders/#disputes"
  },
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-10-09",
    "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"
  }
}