This is a structured service evaluation of Tonewise, an iOS communication-reflection tool grounded in Nonviolent Communication, currently in closed beta. The evaluation is positioned as a Phase 1 feasibility-and-acceptability study, with qualitative reflection as the primary endpoint and quantitative measures as secondary descriptive anchors. It is not a research study under the UK Health Research Authority (HRA) definition: it does not introduce a new intervention, does not aim to generalise beyond the product's user base, and does not require REC or HRA review. It is conducted in line with HCPC standards of conduct, performance and ethics, and the BPS Code of Human Research Ethics. This registration replaces an earlier withdrawn registration (https://osf.io/84cuk/overview), retired before any data collection because the design was substantially restructured following pre-launch review. The substantive changes are: TCCS reduced from 7 to 6 items; weekly battery rewritten as one objective tone-check count, one conditional affect rating, and one optional free-text note; DERS-16 dropped at close-out and retained at baseline only; six qualitative close-out prompts named; recommendation Likert added as primary acceptability endpoint; and analysis plan updated to qualitative-primary, secondary-quantitative-descriptive. No participant has consented and no data exists under the prior version. Approximately 10–15 fellow clinicians (registered psychologists, psychiatrists, and psychotherapists) will be recruited as evaluators. Participants use Tonewise in their own life over six weeks (12 May – 22 June 2026) and contribute data both as users (personal experience) and as clinicians forming a professional view of who in their practice, if anyone, the tool might be useful for. The primary endpoint is qualitative: a six-prompt close-out written reflection (around 600 words total) covering shifts over the six weeks, what helped, what didn't help or got in the way, clinical-view-for-whom, clinical-view-of-concerns, and an open final prompt. The primary acceptability endpoint is a likelihood-to-recommend Likert (0–10) paired with an open rationale, with explicit conflict-of-interest mitigating framing that invites low or critical ratings. Secondary quantitative endpoints are a 6-item study-specific Tonewise Communication Confidence Scale (TCCS, baseline-to-close-out paired); a weekly tone-check count read directly from the in-app home-screen counter (a rolling 7-day window); and a weekly conditional affect rating reflecting how messages that mattered landed after sending. The DERS-16 is administered at baseline only for sample characterisation, not repeated at close-out, given anticipated floor effects in this clinician sample and the qualitative-primary framing. Analysis follows reflexive thematic analysis (Braun & Clarke, 2019) for qualitative data, with a single coder, peer-clinician sense-check, and a reflexive coding journal kept throughout. Quantitative analysis is descriptive, with Wilcoxon signed-rank for paired TCCS change reported alongside explicit small-N caveats. The distribution of recommendation-likelihood ratings will be reported descriptively (median, IQR, full distribution) alongside thematic analysis of the open rationales. Weekly tone-check counts and conditional affect ratings will be reported as time-series, with attention to within-person trajectories rather than between-group means. Conflict of interest: the investigator (Dr Louise Legg, Counselling Psychologist, HCPC, Doctorate) is the founder and sole shareholder of Tonewise Limited (ICO registration ZC127716) and has a direct financial interest in the product's commercial success. This is declared on the participant information sheet, in this registration, and will be declared in all outputs. Outputs comprise an open-access preprint on PsyArXiv (target submission week of 11 August 2026), a plain-language summary returned to each participant before the preprint goes live, and a public summary on usetonewise.com. All findings, including null and negative results, will be reported transparently.