> Earlier version. nomue was an earlier version of SciGround. Its verification capabilities are now part of SciGround; earlier publications, packages, and records keep the nomue name as a historical record.

# nomue documentation (earlier version)

> Archived reference. nomue was Licklider's earlier product, and its verification capabilities are now part of SciGround. These pages describe the named nomue releases and packages only. For current documentation, start at https://www.licklider.ai/docs/index.md.

nomue was Licklider's agent-callable scientific verification product. It moved important scientific decisions and numerical checks out of the research model and into a separate, inspectable capability.

## What was released (September 2026)

- The public @licklider/nomue-verifier package let anyone run nomue verify locally to check a conforming Release 1 Record for independent two-group continuous outcomes under the two-sided Welch two-sample t procedure. It recomputed the covered numerical quantities and returned a machine-readable report of the scoped checks without calling a nomue server after installation.
- [Record verification guide (earlier version)](https://www.licklider.ai/docs/record-verification.md)
- [MCP guide (earlier version)](https://www.licklider.ai/docs/mcp-verification.md): exact local stdio package, configuration, tool-selection boundary, result contract, and independent replay
- [CLI reference (earlier version)](https://www.licklider.ai/docs/cli-reference.md)
- [Examples JSON (earlier version)](https://www.licklider.ai/docs/examples.json)
- Release 1 published source code together with public fixtures, a signed source archive, checksums, a snapshot manifest, detached signatures, and the public release key.

## Limited release, proposals and plans (archived)

- In limited Release 1, approved recipients could submit data and required scientific declarations for a supported Welch calculation, or a claimed result with structured evidence for checking, through authenticated MCP and HTTP. The service returned scoped outcomes, reasons, evidence, and next actions. Public self-registration was not offered, and hosted results were not emitted as public Records for local replay.
- [Hosted limited Release 1](https://www.licklider.ai/news/nomue-welch-limited-release-1/)
- [Making verification results more useful to research agents](https://www.licklider.ai/engineering/making-verification-results-useful-to-agents/): after limited Release 1, nomue development expanded candidate calculation range, improved completion of difficult calculations, clarified what was checked, and added historical-result handling. These were integrated candidates and internal milestones, not a new hosted release or an expansion of public verifier support.
- The exact @licklider/nomue-verifier-mcp@0.2.0-rc.1 release candidate was published on npm and in the official MCP Registry as nomue Record Verifier. It exposed the method-neutral verify_nomue_record tool over local stdio, limited to Release 1 Welch Record verification, delegated to @licklider/nomue-verifier@0.2.1-rc.1, and passed package-path CI across Linux, macOS, and Windows. It required no account, API key, environment variable, or Licklider-hosted service; the first npx launch could download npm dependencies, after which verification ran locally. It supported stdio only: it was not a hosted HTTP endpoint and did not add paired-t, Wilcoxon, Mann–Whitney, method selection, raw-sample calculation, or an overall scientific verdict.
- Release 2 (paired t): the paired-t candidate reached an independently reviewed decision-readiness packet with its numerical and execution evidence, review results, and Release 1 safeguards.
- Release 3 (independent groups and multiple comparisons): public discussion from September 9, 2026, covering 49 catalogued procedures; earliest decision date October 9, 2026.
- Release 4 (balanced two-by-two fixed-factor experiments): public discussion from September 9, 2026; earliest decision date October 9, 2026, and October 18, 2026 for an amendment on strict binary64 comparison and completed indeterminate results.
- Release 5 (shared design and timing evidence across analysis families): public discussion from September 17, 2026; earliest decision date October 17, 2026.
- No decision has been published, and the public nomue Protocol repository records no change after September 28, 2026. Release 1 is unchanged. nomue’s verification work now continues in SciGround.
- Release 3 [discussion record](https://github.com/licklider-ai/nomue-protocol/issues/274); [scope and evidence](https://www.licklider.ai/news/nomue-protocol-release-3-public-discussion/)
- Release 4 [discussion record](https://github.com/licklider-ai/nomue-protocol/issues/261); [scope and evidence](https://www.licklider.ai/news/nomue-protocol-release-4-public-discussion/)
- Release 5 [discussion record](https://github.com/licklider-ai/nomue-protocol/issues/346); [scope and evidence](https://www.licklider.ai/news/nomue-protocol-release-5-public-discussion/)
- The evaluation program examined nomue through decision quality, cost and time, and concrete cases, with explicit comparisons and linked reproduction materials. It published two developer-led preprints on constructed and synthetic Welch workflows and selected cases from the whole-submission study. Results are specific to their tasks, models and configurations; the public archives support reproducing disclosed results, not independently rerunning the private nomue implementations.
- Method families planned as of September 2026: Independent multi-group — SciGround now covers three or more groups (one-way ANOVA with planned comparisons); Paired two-group — SciGround now covers before-and-after comparison (paired t-test); Repeated measures; Factorial and interaction; Nonlinear and dose response; Nonparametric rank-based — SciGround now covers rank comparison of two independent groups (Mann–Whitney U); Categorical outcomes; Correlation and linear models; Survival time-to-event; Count outcomes.
- [Archived nomue roadmap](https://www.licklider.ai/roadmap/)
- [Analyses SciGround supports](https://www.licklider.ai/sciground/analyses/)

## Research finding for call selection

- A statistical method name is not a verification contract. A useful guarantee also depends on the comparison family, error criterion, assumptions, sidedness, balance conditions, and exact procedure variant.
- This source-bounded finding does not add or change Protocol support.

## Documentation

- [What a verification call is (nomue, earlier version)](https://www.licklider.ai/docs/verification-call.md): Archived nomue guide: the shared model nomue used for asking a separate capability to check one bounded part of AI-assisted research.
- [Verify a Release 1 nomue Record (earlier version)](https://www.licklider.ai/docs/record-verification.md): Archived nomue guide: when and how the public local verifier was used for the exact Release 1 Public Draft support target.
- [Use nomue Record verification over MCP (earlier version)](https://www.licklider.ai/docs/mcp-verification.md): Archived nomue guide: the public local stdio server, when its method-neutral Record tool applied, and how a Release 1 Welch result was replayed with the independent verifier.
- [nomue CLI reference (earlier version)](https://www.licklider.ai/docs/cli-reference.md): Archived nomue reference: the exact public command surface shipped by @licklider/nomue-verifier 0.2.1-rc.1, including its help and version behavior.
- [Decision vocabulary (nomue, earlier version)](https://www.licklider.ai/docs/decision-vocabulary.md): Archived nomue guide: how nomue kept execution, clarification, unsupported scope, refusal, failed checks, and unasserted scientific validity separate.
- [nomue examples for agents and implementers (earlier version)](https://www.licklider.ai/docs/examples.md): Archived nomue examples: verifier commands and selection examples showing execute, clarify, unsupported, and bounded interpretation behavior.
- [nomue capability and boundaries (earlier version)](https://www.licklider.ai/docs/limits.md): Archived nomue capability map: what nomue released, what it planned, and how to interpret a successful result, as last updated on September 17, 2026.

## Machine-readable examples

- [Examples JSON (earlier version)](https://www.licklider.ai/docs/examples.json): normalized executable and call-selection examples for nomue, each flagged as the earlier version