# Fix list: Clear Thought 1.5

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/clear-thought, the October 2026 research run, assessed 9 October 2026. Grade F, 36.4 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 Clear Thought 1.5: 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. Reliability, 28 out of 100, up to 14.4 more on the total

Why it scored 28: Scored as local software, since the owner runs it over stdio. An npm package with Node.js 18 or later stated in `engines`. The README's Docker build has no Dockerfile behind it, which issue #17 has reported since 9 September 2025 (18 of 20). No CI workflow in the repository. Two test files exist, and `npm test` calls vitest, which is not among the declared dependencies (3 of 25). Seven open issues and four open pull requests on 9 October 2026. Issues #25 (5 June 2026) and #27 (25 August 2026) report that the stdio server keeps running after the client disconnects, with no reply, and the repository's own report of 24 September 2025 marks a session persistence fault as critical and active (5 of 25). Two versions, 0.2.0 and 0.2.1, with no changelog and no tags, and the server reports 0.0.5 in its handshake (2 of 15). Version 0.2.1, with no stability declaration found (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.

## 2. Security & auth, 39 out of 100, up to 10.7 more on the total

Why it scored 39: No credentials to hold or leak, so the middle band (20). `code_execution` spawns the system Python interpreter with a timeout and no isolation, but only when `allowCodeExecution` is true, and the stdio entry point always starts with false. `notebook_run_cell` is on by default with no switch and runs model-written JavaScript in `node:vm`, which the Node.js documentation says is not a security mechanism. File persistence is off by default (8 of 20). Most output is the model's own input restructured. The server also reads `docs/clear-thought-operations.md` and `../srcbook-examples` relative to the working directory and serves them as resources, and some responses add a line telling the model to run notebook cells with `mcp__ide__executeCode`. No guidance on untrusted content (6 of 15). No per-call log. The `DIAGNOSTIC_ENABLED` variable sets a flag that nothing else in the published build reads (2 of 15). No SECURITY.md, no disclosure contact and no security.txt. GitHub shows no published advisories (1 of 20). The README calls Python execution secure and restricted, and the code shows a plain child process.

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.

## 3. Schema & documentation, 39 out of 100, up to 9.9 more on the total

Why it scored 39: One tool with a JSON Schema generated from zod, read from the source and the published build since we do not run vendor software. Every operation-specific input sits in a free-form `parameters` object and there is no output schema (14 of 25). The README is Markdown and an operations guide is served as an MCP resource. No llms.txt (5 of 10). The 642-character description lists 27 of the 37 operations by name, leaves out the session, notebook, OODA, Ulysses and PDR operations, and does not say when to use the tool or when not to. The README has a short guide to choosing an operation (7 of 20). `operation` is an enum of 37 and `prompt` is required. Nothing inside `parameters` is typed or constrained (5 of 15). The README has eight call examples, two of which (`mdp_planning` and `decision_networks`) name operations the enum rejects. The nine `examples://` resources are looked up relative to the working directory, so they resolve only when the server starts in its package folder. Error responses are not documented (6 of 15). npm version numbers only, with no changelog (2 of 15).

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. Maintenance & community, 4 out of 100, up to 8.4 more on the total

Why it scored 4: The latest npm release, 0.2.1, is dated 18 August 2025, 417 days before this check (0). No release in the last 90 days (0). The last commit on main is dated 24 September 2025. Six of the seven open issues have no comment, the oldest of those from 9 September 2025, and pull request #30, a fix for the disconnect fault, has been open since 25 August 2026 with three others open since October 2025 (2 of 25). Not in the official MCP registry (0). A lockfile and a current MCP SDK range (`^1.17.0`), but no CI, a `test` script whose runner is not declared, and committed build output that issue #28 reports as stale (2 of 10).

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.

## 5. Agent ergonomics, 53 out of 100, up to 7.6 more on the total

Why it scored 53: One tool with a 642-character description, about 160 tokens, and a small schema (25). No way to limit or trim output. Responses are indented JSON, and 32 of the 37 operations append an `initialThought` object from an automatic `sequential_thinking` step. The `advanced` flags that appear to control this are not read by the handler (6 of 20). Invalid arguments return an MCP invalid-parameters error carrying the zod message, and failures inside an operation return `{toolOperation, error, success: false}` with `isError` set. Unknown fields in `parameters` are ignored without a warning, and no error list is documented (9 of 20). No `readOnlyHint`, `destructiveHint` or `idempotentHint`. Each call appends to session state, and the automatic seed step uses one of the 100 thoughts a session allows (3 of 20). Two required arguments with defaults for the rest. `sessionId` and `context` are accepted and dropped (10 of 15).

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.

## 6. Transparency & trust, 36 out of 100, up to 5.6 more on the total

Made of editorial 58, provenance 13.

Why it scored 36: MIT, an OSI licence, on npm and in the repository. The `LICENSE` file's copyright line names Anthropic, PBC, carried over from the MCP reference servers, and not the maintainer (28 of 30). No written statement on data handling. The source shows state held in memory, optional file persistence to `.ct-data` that is off by default, and no network calls in the published build. The main handler is one file of about 2,900 lines, so the reading takes effort (15 of 30). No deprecation policy or dated notices, and the earlier `@waldzellai/clear-thought` package is not marked deprecated on npm (0 of 20). The published build has no telemetry. Main adds Shinzo instrumentation that is off by default and needs a token, described in the configuration schema and a commit message and not in the README (15 of 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):

- Legal entity named: not found (0 of 20)
- Domain age: github.com/waldzellai, no registry record we could read (0 of 15)
- Status page: not found (0 of 10)
- Changelog: not found (0 of 10)
- security.txt: not found (0 of 10)

## 7. Payments & pricing, 60 out of 100, up to 5 more on the total

Why it scored 60: Free, self-hosted software with 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.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- Seen 2026-10-09. The README and package description state things the code does not support. The package description says 38 operations and the enum has 37, the README shows `mdp_planning` and `decision_networks` calls that the schema rejects, it gives a Docker build with no Dockerfile in the repository (issue #17, open since 9 September 2025), and it describes Python execution as secure and restricted where the code spawns the system interpreter with only a timeout. The README dates from 24 September 2025 and still reads this way, so the smallest deduction applies (https://github.com/waldzellai/clearthought-onepointfive/blob/main/README.md).

## 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: the Smithery listing at smithery.ai/servers/waldzellai/clear-thought-onepointfive. The page is drawn by script and showed our reader only navigation, so whether a hosted endpoint is live, and its use count, are unknown.
- unchecked: the bodies and comment threads of the open issues and pull requests. Only the list pages were read, so the disconnect fault is reported from issue titles and the open fix in pull request #30.
- unchecked: behaviour at run time. Nothing was installed or run, so the operation count, the ignored arguments and the working-directory lookups come from reading the source and the published build.
- Whether the session persistence fault in the repository's report of 24 September 2025 affects stdio use. The report describes calls that each receive a new session, which matches the HTTP deployment and not a single stdio process.
- Who Waldzell AI is in law. No legal entity, website, terms or privacy policy was found.
- The lead described the product correctly. Its package description says 38 operations, and the schema has 37.

## Weaknesses

- No npm release since 0.2.1 on 18 August 2025 and no commit on main since 24 September 2025
- Six of seven open issues have no reply, among them two reports that the stdio server keeps running after the client disconnects
- Operation-specific input is a free-form `parameters` object, and `sessionId`, `context` and `advanced` are accepted but never read
- `notebook_run_cell` runs model-written JavaScript in `node:vm`, which the Node.js documentation says is not a security mechanism
- The README shows `mdp_planning` and `decision_networks` calls the schema rejects, and a Docker build with no Dockerfile in the repository

## 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.

- Call `clear_thought` with `operation` and `prompt`. Put everything else in `parameters`, using the field names in the `guide://clear-thought-operations` resource
- Do not call `mdp_planning` or `decision_networks`. The README shows them, and the tool's operation enum rejects both
- Expect an extra `initialThought` object on 32 of the 37 operations, and each one uses a slot of the 100-thought session limit
- Stop the server process yourself when the session ends. Open issues report that it does not exit when the client disconnects
- Treat `notebook_run_cell` as unsandboxed code execution on the host, and do not pass it code from untrusted content

## 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.