·      The original Dokwe drilling database was in the form of a Microsoft access database. The Dokwe drillhole database included all drilling prior to 2023. This data was imported to MXDeposit. All data collected during the 2023-2024 due diligence drilling programme was added directly to MXDeposit.

·      The QA/QC for the various drilling campaigns was reviewed and deemed suitable for the results to be used in a mineral resource estimate. The Dokwe drillhole database was checked for duplicates, overlapping and missing intervals on import into Leapfrog, whilst all fields were checked for spurious or out-of-range values. Any errors were corrected prior to modelling.

·      As Competent Person for the Mineral Resources presented in this Report, Zack van Coller conducted site visits in November 2023, March 2024 and January 2025. Drillhole collar positions were confirmed, and diamond drill core was inspected in the core yard. It was confirmed that the mineralisation is disseminated and not related to a distinct lithology or structural feature. Varying degrees of deformation were observed in association with more mineralisation. The CP was present for some of the due diligence diamond drilling programme and sampling.

·      There is a high level of confidence in the geological interpretation, the deposit is well sampled, and the density of data allows for a suitable interpretation of the grade distribution.

·      A sub-selection of the original drillhole logs and laboratory assay certificates were compared to the final Dokwe drillhole database. The CP was present during the logging of the 2023-2024 drillholes, and again in January 2025 for detailed review of high-grade zones.

DOKWE NORTH

·      Digital Mining Services completed an MRE as an updated statement in January 2020. This estimate was largely focused around two explicitly modelled grade wireframes at 0.5g/t and 0.2g/t Au. Only composites from within the 0.5g/t Au grade shell were considered in that estimate, whilst for the 0.2g/t Au shell estimate, the remainder of the composites (excluding composites from within the 0.5g/t Au shell) were used for interpolation of that shell. This resulted in a distinct grade boundary between the two shell estimates. This "hard" boundary in grades may not necessarily be evident in the distribution of grades present in the drillhole data.

·      The estimation and Mineral Resource categorisation methodologies between the January 2020 estimate and the estimate presented in the 2022 PFS, are significantly different, and have resulted in significant differences in terms of both volume and grade for all Mineral Resource categories.

·      The 2022 Minxcon Mineral Resource estimate presented in the PFS represents a broader implicit grade shell (at 0.2g/t Au) estimate, and an internal 0.7g/t Au sub-domain, and would result in larger volumes and lower average grades than the previous estimation methodology.

·      2024 Estimation: A lithological model was used to constrain the estimation of grade into the block model, with gold estimated separately into each lithology domain. Grade clamping was applied (instead of a top-cut) so as to preserve the high grades, but minimise the distance the grade can be spread. This was a 50, 20 and 10g/t Au clamp for passes 1, 2 and 3, respectively.

·      2025 Estimation: The biggest change since the previous resource work is the new mineralisation model, which constrains mineralisation in three geologically and structurally defined domains, leading to a far more robust mineralisation interpretation model. A high-grade interpolant model at a modelling cut-off of 0.7g/t Au was used to isolate high-grade data through the deposit. High-grade within this particular domain was capped to 200g/t Au and then further constrained with the application of an Outlier Restrictor, clamping of high grades were set to 100g/t Au and limited to 5-10 meter extrapolation. High grades were also evaluated during the variography analysis stage of the estimation, where further capping was applied to improve variogram correlation.

DOKWE CENTRAL

·      DMS completed an MRE in 2011. The 2011 estimation domains were manually constructed wireframes base on three vertically dipping bodies. The 2011 MRE was evaluated to be very conservative with wireframe boundaries being very restrictive in joining clusters of mineralised composites.

·      2024 and 2025 Ariana Estimates: A lithological model was used to constrain the estimation of grade into the block model, with gold estimated separately into each lithology domain. A top-cut of 30g/t Au was applied to minimise the distance the grade can be spread. The estimate was completed in two search passes of 10x20x40m and 20x40x80m.

·      The 2024 and 2025 estimation domains are based on a lithological model between two primary geological packages (hanging wall biotite schists (mineralised) and footwall sedimentary gritstones/conglomerates (barren)). Separating the two geological packages is a significant east-west trending fault (very similar to the 2011 interpretations). Mineralisation was restricted from extrapolating across to the "barren" sedimentary units. Within the biotite mineralisation package search ellipse inputs were used from the 2011 MRE to re-establish and validate mineralisation continuity in the new model.

·      Interpolation modelling with support from pXRF data from eight diamond drill holes, as well as additional new drill holes drilled at the deposit since 2011 (mainly two new holes drilled by Ariana in 2023), is thought to have made more representative iso-surface volumes as an update to mineralisation interpretations.

·      In 2025 the Dokwe Central geological model was further refined to include a second major N-S fault zone, adding restrictions to the model's mineralisation extrapolation. As a result, Inferred resources were reduced to better reflect the influence of the bounding structures.

·    The Dokwe North orebody is 780m along strike, 470m across strike (across the thickest portion of the deposit), and the depth from the surface is between 42m and 320m.

·    The Dokwe Central orebody is 260m by 200m across and the depth from the surface is between 25m and 350m.

·      Leapfrog Geo 2023.2 software was used to construct the geological wireframes/mineralised halos, while Leapfrog Edge 2023.2 was used to conduct statistical and geostatistical analyses and generate the estimated block model.

·      No assumptions were made in terms of selective mining units with respect to the cell size selected.

·      No assumptions were made regarding correlation between variables.

·      Several data-model reconciliations were performed. Firstly, a visual inspection of drillhole composite values with respect to the estimated block model was completed.

DOKWE NORTH

·      An Ordinary Kriging estimate was completed. Swath plots indicate a good correlation between drilling data and estimated block grades.

·      A variography study was completed for each mineralisation domain used in the estimation. All domains produced good variograms. To reduce bias and improve the variography, a top-cut analysis was completed during the variography study. This is separate from the additional top-cut and Outlier Restriction analysis completed later in the estimation. 

·      In summary, the various validations and reconciliation techniques demonstrate that the block model estimates show a good correlation between interpolation methods and with the informing composites. Furthermore, the estimation quality and low conditional bias parameters appear to indicate that the estimation technique has provided an acceptable estimate without excessive smoothing.

·      A 47 degree rotated sub-block model was established using block sizes determined to be optimal for the dataset (50m collar spacing) and wireframe geometry, with sub-blocks triggered by mineralisation boundaries. For Dokwe North the parent blocks are 10m x 20m x 10m (X,Y,Z), with sub-blocks of 5m x 5m x 5m (X,Y,Z).

·      Three domains were modelled and estimated.

1) Zones of intense ductile deformation logged through the deposit as foliation intensity; as a means to map out the primary extent of a mineralising shear zone (Shear Zone Domain)

2) A sub-domain within the Shear Zone Domain, which is statistically defined to be a sub-population of "high-grade" data, which uses a 0.7g/t Au interpolant shell with extrapolation to a maximum distance of 150m (High-grade Domain)

3) A wider interpolant, modelling the maximum footprint of mineralisation in all peripheral areas away from the main shear zone; modelled at a 0.1g/t Au model cut-off with a 150m extrapolation (Low-grade Domain).

Table below: Kriging ellipse input for all domains:

Table below from top to bottom, Low-grade domain, High-grade domain and wider Shear Zone domain:

Summary of Variography inputs used for the High Grade Domain:

Summary of Variography inputs used for the Low Grade Domain:

Summary of Variography inputs used for the Shear Zone Domain:

DOKWE CENTRAL

·      An IDW2, Inverse Distance estimation, was used for Dokwe Central, as this was deemed most appropriate for the nature of the deposit, and statistical outputs An Ordinary Kriging estimate was also completed as a means to check the IDW2 estimation. This produced similar grades and tonnages to the IDW2 method. However, suitable variograms were not defined. 

·      In summary, the various validations and reconciliation techniques demonstrate that the block model estimates show a good correlation between various interpolation methods and with the informing composites. Furthermore, the estimation quality and conditional bias parameters appear to indicate that the estimation technique has provided an acceptable estimate without excessive smoothing.

·      An orthogonal non-rotated block model was established using block sizes determined to be optimal for the dataset (30m collar spacing) and wireframe geometry. For Dokwe Central this was 10m x 10m x 5m (X,Y,Z).

·    The density is based on water displacement measurements which give the dry rock mass.

·    Using the mining factors refined from the 2022 pre-feasibility study on Dokwe, the actual cut-off grade that was determined was 0.26g/t Au. However, the CP have opted for a higher cut-off grade of 0.6g/t Au for the Mineral Resource cut-off grade.

·      Table below: Optimisation inputs used for the 2024/2025 MRE work for Dokwe North and Dokwe Central:

DOKWE NORTH

·      Measured, Indicated and Inferred Mineral Resources have been stated within a revised 2025 optimisation resource pit shell based on a $2750 gold price.

DOKWE CENTRAL

·      The Mineral Resource has been declared both within a resource open pit shell and no pit shell. The pit shell is a preliminary optimisation carried out in 2022 by Axe Valley Mining Consultants Limited.

·      Work was completed post-resource update to review the optimisation outputs based on a revised gold oz price ($2,000/oz and $2,500/oz).

DOKWE NORTH

·      The processing of oxide material is envisaged to be done using conventional CIL processing as limited preg-robbing properties were identified. The transitional and sulphide material will likely be processed through flotation with high-intensity leaching (CIP) of the flotation concentrate.

·      The table below shows the metal recoveries determined from metallurgical test work. In summary, of the total gold content, 25.9% is recovered by gravity, with 61.35% by flotation and intense leach - giving a total recovery of 87.35%.

Table Below: Dokwe North MRE in 2025 optimisation pit (at $2750 gold price), by oxidation state of mineralisation based on detailed drill core logging. Reporting at a 0.3g/t Au cut-off.

DOKWE CENTRAL

·      Detailed metallurgical testwork has not been completed on Dokwe Central but some samples were included in early composites. There are no indications that gold recovery will be problematic, particularly following the results of the detectORE^TM analysis undertaken on the due diligence drilling core.

·      Oxide and Transitional components of the sulphide zone dominant Dokwe Central mineralisation have been defined and make up a small proportion of the mineralisation. Since Dokwe Central will be processed alongside Dokwe North, it is assumed processing of oxide material will be done using conventional CIL processing. The transitional and sulphide material will likely be processed through flotation with high-intensity leaching (CIP) of the flotation concentrate. 

Table Below: Dokwe Central MRE in 2025 optimisation pit (at US$2750/oz gold price), by oxidation state of mineralisation based on detailed drill core logging. Reported at a 0.3g/t Au cut-off.

·    No environmental factors or assumptions were applied to this Mineral Resource Estimation.

·      Specific gravity measurements have been collected during the resource drilling at Dokwe. A programme of sampling across strike on 3 lines was undertaken on the pre-2019 drilling resulting in 100 density measurements. During the 2019 drilling campaign, 6 drillholes on 6 lines in the south-eastern portion of the project were sampled for density in a much more comprehensive programme, with 327 measurements being taken.

·      On average, 18cm core samples were measured. The samples were weighed in the air, and then weighed in water, the SG was calculated, by dividing the weight of the sample in the air by the weight of the sample in the water. Samples were sealed with grease to prevent water ingress and ensure that they any porosity was taken into account. The table below presents average SG for different oxidation type.

·      A Specific Gravity ("SG") estimation model was established for Dokwe North. 475 SG readings from 22 drillholes were coded into the final block model. This data includes 158 verification samples taken by Ariana during the company's due diligence review. Average SG measurements across Dokwe North range from 2.71g/cm³ in the oxide zone, 2.76g/cm³ in the transitional zone to 2.81g/cm³ in the sulphide zone.

·      No historical SG data exists for Dokwe Central. However, Ariana acquired 92 measurements through the deposit profile from two drill holes completed during its 2023 due diligence drilling programme. The average SG for these 92 measurements is 2.69g/cm³. This was applied to the Dokwe Central model as a representative flat rate.

DOKWE NORTH

·      The Mineral Resource is classified and reported in accordance with the 2012 JORC Code as Measured, Indicated, and Inferred. The classification is determined based on kriging efficiency and distance from drilling. These are given in more detail under the section "Estimation and modelling techniques".

·      Measured Mineral Resources have been defined using a search pass ellipse with a search diameter of 30mx15mx10m, as well as a review of kriging efficiency and slope of regression statistics. From this, a volume was built to capture the most appropriate volume for the highest confidence-spaced data. 

·      Indicated Mineral Resources have been defined using a search pass ellipse with a search diameter of 60mx30mx20m, as well as a review of kriging efficiency and slope of regression statistics. From this, a volume was built to capture the most appropriate volume for the next highest confidence-spaced data. 

·      Inferred Mineral Resources have been defined using a search pass ellipse with a search diameter of 180mx60mx40m, as well as a review of kriging efficiency and slope of regression statistics. From this, the remaining available volume within the mineralisation model was filled to maximise the expanse of mineralisation extrapolation to a maximum distance of 180m.

DOKWE CENTRAL

·      The Mineral Resource is classified and reported in accordance with the 2012 JORC Code as Measured, Indicated and Inferred. The classification is determined based on search pass spacing, with increasing confidence with proximity to drill holes. These are given in more detail under section "Estimation and modelling techniques".

·      Measured Mineral Resources have not been defined.

·      Indicated Mineral Resources have been defined by Pass 1 (up to 10m x 20m x 40m).

·      Inferred Mineral Resources have been defined in areas beyond the Indicated search radius to the limits of the resource wireframes in Pass 2 (up to 20m x 40m x 80m).

·    Internal reviews of the Mineral Resource estimate were completed.

·      Several data-model reconciliations were performed. Firstly, a visual inspection of drillhole composite values with respect to the estimated block model was completed. Visually there is a good correlation between the estimated ordinary kriging gold values and the composite gold values, and the raw assay data.

·      Basic statistics have been compiled comparing the model estimates and composites.

·      In summary, the various validations and reconciliation techniques demonstrate that the block model estimates show a good correlation between various interpolation methods and with the informing composites. Furthermore, the estimation quality and conditional bias parameters appear to indicate that

the estimation technique has provided an acceptable estimate without excessive smoothing.

·      Overall wider block distribution accuracy is considered acceptable as evidenced by direct drillhole verses block model checks, ensuring acceptable localised accuracy.

·      Accuracy of the estimate relative to production data cannot be ascertained at this point as there is no production.