153 lines
38 KiB
JSON
153 lines
38 KiB
JSON
{"prompt":"Design a reproducible scientific-workflow platform for simulations, parameter sweeps, checkpoints, datasets, notebooks, provenance, and published results. Cover scheduler integration, immutable environments, retries, quotas, lineage, regional storage, migration from shell scripts, and rollout milestones. Deliver an RFC only.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.95,"slice":"core","lang":"en"}
|
|
{"prompt":"Before implementing ensemble weather forecasts, outline data ingestion, member scheduling, checkpointing, grid consistency, aggregation, late members, publication, validation, and phased rollout.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.94,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Add NetCDF chunk inspection to `climatectl`","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.64,"slice":"core","lang":"en"}
|
|
{"prompt":"Simulation failure output:\n\n```text\nrun=atm_881 model=wrf-5.0 nodes=64 ranks=4096\nstep=1440 simulated_time=2026-07-30T12:00Z\ncheckpoint generation=18 begin\nrank 0 manifest uploaded parts=4095 expected=4096\nrank 2917 object upload timeout after=120s\nscheduler preempts allocation at=12:01:59\nretry resumes generation=18 because manifest state=writing\nmissing rank file is treated as zero-length optional boundary partition\nsimulation advances and pressure field develops NaNs near partition edge\nvalidation checks global file count only after merged output\n```\n\nDetermine why an incomplete distributed checkpoint is considered resumable and why the missing rank is not rejected before computation continues.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.89,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Build the SwiftUI experiment-run screen with status, parameters, resource usage, charts, artifacts, cancellation, and VoiceOver summaries.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.65,"slice":"core","lang":"en"}
|
|
{"prompt":"Plan and implement a provenance service for scientific datasets and model runs. Begin with a design covering content identity, transformations, environments, external sources, mutable metadata, citations, access, retention, verification, and migration. Then add lineage APIs, signed manifests, query support, and conformance tests. Treat reproducibility semantics as primary.","purpose":"planning","secondary":"backendImpl","mixed":true,"difficulty":0.97,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Set `MAX_ENSEMBLE_MEMBERS` to 100","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.07,"slice":"core","lang":"en"}
|
|
{"prompt":"Extract duplicated coordinate-axis normalization from forecast and observation readers while preserving values.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.63,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement `/internal/jobs/capacity` with partitions, nodes, accelerators, queued demand, and reservations.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.53,"slice":"boundary","lang":"en"}
|
|
{"prompt":"make the charts feel scientific","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.4,"slice":"vague-eval","lang":"en"}
|
|
{"prompt":"Find why CUDA kernels produce stable results on A100 nodes but occasional infinities on H200 nodes.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.83,"slice":"core","lang":"en"}
|
|
{"prompt":"Explain how `GridReprojector` handles antimeridian cells, polar singularities, and missing values. No edits.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.68,"slice":"core","lang":"en"}
|
|
{"prompt":"Research release notes source:\n\n```text\nFeature: citable dataset snapshots\nResearchers create an immutable snapshot from one dataset revision\nSnapshot includes schema, files, checksums, licenses, and selected metadata\nA DOI can be requested after institutional review\nMetadata corrections after publication create a new descriptive revision without changing file digests\nEmbargoed snapshots are visible only to authorized collaborators until release time\nDeleting a source dataset does not remove a published snapshot\nStorage charges remain with the owning project\n```\n\nTurn this into researcher-facing release notes and a publication guide. Clearly distinguish data immutability from metadata correction.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.46,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Move the chart legend below the color scale","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.13,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Write an operator guide for replaying failed observation ingests without duplicating station intervals.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.42,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement dataset-snapshot endpoints with file selection, checksums, and authorization, then document lifecycle, examples, and errors.","purpose":"backendImpl","secondary":"writing","mixed":true,"difficulty":0.69,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Map out a migration from shared NFS scratch space to object-backed workflow storage, including performance, consistency, checkpoints, compatibility, and rollback.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.91,"slice":"core","lang":"en"}
|
|
{"prompt":"Rename `gridNo` to `gridNumber` in `ForecastMapper.py`.","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.04,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Teach the batch scheduler to reserve topology-aware GPU groups and release incomplete allocations.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.72,"slice":"core","lang":"en"}
|
|
{"prompt":"Polish the dataset summary card","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.28,"slice":"core","lang":"en"}
|
|
{"prompt":"Bump `xarray` to `2026.7.0`","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.07,"slice":"core","lang":"en"}
|
|
{"prompt":"Rename `MasterDataset` to `PrimaryDataset` across APIs, events, dashboards, and tests while preserving public identifiers.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.64,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Split `WorkflowRunner` into planning, staging, execution, checkpointing, and publication components without changing outcomes.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.81,"slice":"core","lang":"en"}
|
|
{"prompt":"Implementation ticket:\n\n```text\nSCI-611 — station observation corrections\nMeasurements keyed by station, variable, level, and observation time\nProviders may send preliminary, corrected, or final values\nCorrection messages reference provider record ID but may arrive out of order\nFinal values cannot be replaced by preliminary updates\nUnit conversion occurs before quality-control rules\nRaw payload and transformation version must remain available\nDuplicate provider messages are common\nExpose preliminary, corrected, rejected-stale, converted, and quality-failed metrics\n```\n\nImplement the version model, ingest commands, ordering, conversion, quality checks, provenance, metrics, and tests.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.84,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Investigate why the Android field app assigns a sensor reading to the previously selected sampling site.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.69,"slice":"core","lang":"en"}
|
|
{"prompt":"Create the provenance graph UI and add a short legend for sources, transforms, environments, snapshots, and publications.","purpose":"frontendImpl","secondary":"writing","mixed":true,"difficulty":0.68,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Why does the workflow service retain failed checkpoints for 30 days? Explain current debugging behavior.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.45,"slice":"core","lang":"en"}
|
|
{"prompt":"Rewrite the local setup for running Slurm, MinIO, JupyterHub, and synthetic climate jobs.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.36,"slice":"core","lang":"en"}
|
|
{"prompt":"This reprojection test fails after upgrading PROJ:\n\n```text\nsource CRS=EPSG:4326 target CRS=EPSG:3413\ninput lon=-45 lat=90\nexpected x=0 y=0\nactual x=0 y=-1.2e-9\nfixture uses exact equality\nbulk test reports 812 changed cells near pole\nmaximum absolute difference=3.7e-8m\noutput is later quantized to 0.01m\nchecksum fixtures compare pre-quantized floats\n```\n\nDetermine whether the library change is a real projection regression or harmless floating-point drift, and identify which validation layer uses the wrong tolerance.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.69,"slice":"boundary","lang":"en"}
|
|
{"prompt":"finish the workflow service","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.48,"slice":"vague-eval","lang":"en"}
|
|
{"prompt":"Propose an architecture for real-time satellite-data processing, including downlink ingestion, calibration, tiling, cloud masking, provenance, late passes, and publication.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.95,"slice":"core","lang":"en"}
|
|
{"prompt":"Add GRIB2 bitmap decoding to `weatherdump`","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.68,"slice":"core","lang":"en"}
|
|
{"prompt":"Review excerpt:\n\n```diff\n def publish_result(run_id):\n- manifest = verify_all_outputs(run_id)\n- if not manifest.valid:\n- raise InvalidResult(manifest.errors)\n- registry.publish(run_id, manifest.digest)\n+ future = executor.submit(verify_all_outputs, run_id)\n+ registry.publish(run_id, 'pending')\n+ future.add_done_callback(log_verification)\n```\n\nThe author says verification delays researchers. Review publication semantics, invalid results, immutability, citations, audit, and asynchronous failure. No patch requested.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.75,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Schreibe eine kurze Referenz für die bestehenden Workflow-Status und ihre Auswirkungen auf Artefakte und Wiederaufnahme. Kein Code.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.37,"slice":"boundary","lang":"de"}
|
|
{"prompt":"Build the desktop geospatial raster viewer with layers, projections, color maps, value inspection, and keyboard controls.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.72,"slice":"core","lang":"en"}
|
|
{"prompt":"Review how dataset checksums are calculated and verified, then turn the findings into a publication guide with examples.","purpose":"writing","secondary":"review","mixed":true,"difficulty":0.6,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Set the checkpoint interval to 30 minutes","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.07,"slice":"core","lang":"en"}
|
|
{"prompt":"Benenne `RunCtx` repositoryweit in `RunContext` um; JSON-Felder und Verhalten bleiben unverändert.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.57,"slice":"boundary","lang":"de"}
|
|
{"prompt":"Consolidate missing-value normalization across NetCDF and GRIB readers without changing arrays.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.66,"slice":"core","lang":"en"}
|
|
{"prompt":"Diagnose the duplicate dataset alert","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.43,"slice":"core","lang":"en"}
|
|
{"prompt":"Explain how the scheduler handles reservations when one node in a tightly coupled job fails health checks.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.63,"slice":"core","lang":"en"}
|
|
{"prompt":"Create API documentation for the existing workflow-cancellation endpoint, including states, cleanup, and errors.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.37,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Plan a migration from ad hoc Conda environments to signed OCI research environments, covering reproducibility, GPUs, secrets, storage, and rollback.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.9,"slice":"core","lang":"en"}
|
|
{"prompt":"Visualization ticket:\n\n```text\nPage: ensemble forecast explorer\nControls: variable, pressure level, forecast time, member, aggregation\nMap: mean field with optional spread overlay\nChart: point forecast showing all members plus percentiles\nTime controls must retain selected geographic point\nMissing members shown explicitly; do not treat as zero\nColor scales have units and accessible text alternatives\nLarge datasets load progressively without resetting zoom\nComparison mode pins two model runs with matched valid times\n```\n\nImplement the responsive explorer and component tests using existing forecast APIs.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.76,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Implement `POST /v1/runs/{id}/resume` with checkpoint validation, environment matching, and audit events.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.66,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Remove the extra unit from the axis label","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.03,"slice":"core","lang":"en"}
|
|
{"prompt":"Fix the clipped map tooltip and extract its duplicated viewport inset into shared styles.","purpose":"quickFix","secondary":"refactor","mixed":true,"difficulty":0.36,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Replace scattered run booleans with one explicit workflow lifecycle state while preserving transitions and UI.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.67,"slice":"core","lang":"en"}
|
|
{"prompt":"Cambia `gridRev` por `gridRevision` solo en `ForecastState.swift`; no cambies la lógica.","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.06,"slice":"boundary","lang":"es"}
|
|
{"prompt":"Investigate why the research portal briefly shows another project's private dataset after workspace switching.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.79,"slice":"core","lang":"en"}
|
|
{"prompt":"Summarize the current dataset-publication lifecycle for research-integrity reviewers using code and tests.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.44,"slice":"core","lang":"en"}
|
|
{"prompt":"Develop a capacity roadmap for storing and serving 200 PB of climate data. Include chunking, tiers, catalogs, query patterns, integrity, and cost gates.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.95,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement idempotent model-output ingestion with run generations and correction support.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.71,"slice":"core","lang":"en"}
|
|
{"prompt":"Create the Compose field-observation form with site, variables, units, quality flags, offline queue, and TalkBack support.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.66,"slice":"core","lang":"en"}
|
|
{"prompt":"Fix `preciptation` in `variables.yaml`","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.02,"slice":"core","lang":"en"}
|
|
{"prompt":"Merge duplicate spatial-bounds types used by catalog and visualization while preserving serialized values.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.64,"slice":"core","lang":"en"}
|
|
{"prompt":"Forecast publication incident:\n\n```text\nrun=gfs_20260730_00 member=control\noutput valid_times=0..384h\npublication manifest generation=771 references 129 files\nobject-store list was eventually consistent and omitted hour=240\nmanifest validator expected at least 120 files, passed\ncatalog published run complete\nfile hour=240 appeared 14s later but was not in immutable manifest\nusers receive gap at one forecast time\nretry refuses new manifest because run marked published\n```\n\nDetermine why count-based validation and eventual listing produced an incomplete immutable publication and how completeness should have been established.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.82,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Rename `longtitude` to `longitude` in the unreleased schema.","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.06,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Assess whether shared notebook environments can expose another project's mounted datasets or credentials. Review isolation only.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.73,"slice":"core","lang":"en"}
|
|
{"prompt":"Restructure the climate SDK into catalog, datasets, grids, forecasts, and provenance, then write a maintainer note. Preserve APIs.","purpose":"refactor","secondary":"writing","mixed":true,"difficulty":0.75,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Plan the retirement of legacy FTP observation delivery, covering providers, parity, acknowledgements, migration, and removal gates.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.83,"slice":"core","lang":"en"}
|
|
{"prompt":"Add immutable provenance chains to calibrated satellite scenes.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.7,"slice":"core","lang":"en"}
|
|
{"prompt":"Build the tablet sampling-route view with sites, offline maps, collection state, navigation, and accessible list fallback.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.67,"slice":"core","lang":"en"}
|
|
{"prompt":"do the forecast page","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.4,"slice":"vague-eval","lang":"en"}
|
|
{"prompt":"Replace `DataRef` with `DatasetReference` across packages while preserving serialized fields.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.56,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Diagnose why published figures use an older dataset revision, implement the verified snapshot-binding fix, and add a researcher note.","purpose":"debugging","secondary":"writing","mixed":true,"difficulty":0.79,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Why does the workflow service retain two previous container images after environment updates? Explain rollback behavior.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.46,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Track down the memory leak in raster tilers after processing high-band-count scenes.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.73,"slice":"core","lang":"en"}
|
|
{"prompt":"Write release notes for parameter-sweep comparisons, including aggregation, missing runs, exports, and limits.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.36,"slice":"core","lang":"en"}
|
|
{"prompt":"Workflow checkpoint notes:\n\n```text\nCurrent checkpoints are opaque directories written by each model\nNew platform needs portable checkpoint manifests\nFiles may be shared across ranks through hard links\nA checkpoint is valid only when all ranks complete a barrier\nEnvironment, executable, configuration, and input dataset revisions must be captured\nUsers may resume with more nodes only if model declares repartition support\nPartial checkpoints should remain available for debugging but never resume automatically\nOld model versions emit no schema version\n```\n\nProduce an implementation-free format and migration plan covering completeness, identity, portability, validation, partial artifacts, compatibility, and rollout.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.92,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Give the run list clearer state grouping","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.31,"slice":"core","lang":"en"}
|
|
{"prompt":"Turn the current data-embargo behavior into researcher documentation, then review whether release timing and access claims match the implementation.","purpose":"writing","secondary":"review","mixed":true,"difficulty":0.68,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Outline a phased architecture for moving workflow metadata from MongoDB to Postgres, without implementing migrations.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.82,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Implement version-aware station metadata history with effective dates and correction links.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.7,"slice":"core","lang":"en"}
|
|
{"prompt":"Create an operator guide for diagnosing Slurm jobs stuck in completing state.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.4,"slice":"core","lang":"en"}
|
|
{"prompt":"Move unit-conversion selection out of `ObservationController` while preserving values.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.59,"slice":"core","lang":"en"}
|
|
{"prompt":"Investigate why revoked collaborators retain access through one dataset tile cache.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.71,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Add generation-fenced model-configuration publication with validation and rollback.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.65,"slice":"core","lang":"en"}
|
|
{"prompt":"Set the plot export DPI to 300","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.07,"slice":"core","lang":"en"}
|
|
{"prompt":"Create the responsive dataset-browser with facets, previews, licenses, lineage, downloads, and accessible tables.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.66,"slice":"core","lang":"en"}
|
|
{"prompt":"Corrige apenas `temprature` para `temperature` no novo gráfico.","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.03,"slice":"boundary","lang":"pt"}
|
|
{"prompt":"Review the notebook-export endpoint for embedded credentials, private file leakage, unsafe HTML, huge outputs, and stale authorization.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.74,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement workflow-run history and write its API guide. Include inputs, environment, checkpoints, outputs, pagination, and permissions.","purpose":"backendImpl","secondary":"writing","mixed":true,"difficulty":0.71,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Design a migration from static station files to a versioned geospatial registry, including identity, moves, corrections, compatibility, and rollout.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.89,"slice":"core","lang":"en"}
|
|
{"prompt":"Add resumable multi-file dataset uploads with checksums and atomic publication.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.71,"slice":"core","lang":"en"}
|
|
{"prompt":"Build the desktop ensemble-comparison view with synchronized maps, distributions, valid times, and export.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.7,"slice":"core","lang":"en"}
|
|
{"prompt":"continue the checkpoint bug","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.56,"slice":"vague-eval","lang":"en"}
|
|
{"prompt":"Extract coordinate-reference normalization into one shared module while preserving identifiers.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.58,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Investigate why observation alerts use an outdated station elevation, implement the verified metadata-version fix, and update operator docs.","purpose":"debugging","secondary":"writing","mixed":true,"difficulty":0.77,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Assess the proposed public forecast cache for embargoes, licensing, stale corrections, regional policy, and invalidation.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.7,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Find why one workflow partition retries an impossible resource request forever.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.64,"slice":"core","lang":"en"}
|
|
{"prompt":"Draft a migration guide for researchers moving from local scripts to managed workflows.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.39,"slice":"core","lang":"en"}
|
|
{"prompt":"Numerical validation report:\n\n```text\nmodel=ocean_v12 baseline compiler=gcc-14 new compiler=gcc-15\nrun length=30 days\nsea_surface_height max_abs_diff=1.8e-7 day=1\nmax_abs_diff=2.1e-3 day=30\nenergy conservation baseline drift=0.0012% new drift=0.0013%\nbitwise checksum mismatch from first timestep\nstatistical climatology remains within tolerance\nvalidation pipeline requires bitwise equality for restart files\nnew compiler enables fused multiply-add in one kernel\n```\n\nDetermine whether the divergence is an unacceptable reproducibility regression or expected chaotic amplification, and which validation criteria should govern restart compatibility versus scientific equivalence.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.84,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Add valid-time text below the forecast title","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.24,"slice":"core","lang":"en"}
|
|
{"prompt":"Restructure the observation pipeline and document its provider interface. Preserve data while separating decode, normalize, quality control, correction, and publication.","purpose":"refactor","secondary":"writing","mixed":true,"difficulty":0.81,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Architect a multi-region scientific-data platform with dataset ownership, provenance, publication, compute locality, and migration milestones.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.96,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Implement signed workflow manifests with environment, inputs, parameters, and output digests.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.72,"slice":"core","lang":"en"}
|
|
{"prompt":"Write an onboarding guide for adding a new forecast model, including grids, variables, workflows, validation, and publication.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.47,"slice":"core","lang":"en"}
|
|
{"prompt":"Consolidate scientific-time formatting across catalog and exports without changing output.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.55,"slice":"core","lang":"en"}
|
|
{"prompt":"Investigate why withdrawn datasets remain in public catalog autocomplete.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.61,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Add transactional dataset ownership transfer with permission remapping and provenance.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.72,"slice":"core","lang":"en"}
|
|
{"prompt":"Pin GDAL to `3.11.3`","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.06,"slice":"core","lang":"en"}
|
|
{"prompt":"Build the mobile field-sample history with sites, variables, quality, sync state, and accessible summaries.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.59,"slice":"core","lang":"en"}
|
|
{"prompt":"Remplace `Forcast` par `Forecast` uniquement dans le nouveau titre.","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.03,"slice":"boundary","lang":"fr"}
|
|
{"prompt":"Inspect the workflow-definition import path for command injection, unsafe mounts, secret exposure, huge DAGs, and tenant confusion.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.78,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement dataset-publication history and write its API documentation. Include revisions, reviews, embargoes, DOI status, pagination, and permissions.","purpose":"backendImpl","secondary":"writing","mixed":true,"difficulty":0.71,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Develop a roadmap for replacing manually curated quality rules with versioned executable policies, including simulation, approval, and rollout.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.88,"slice":"core","lang":"en"}
|
|
{"prompt":"Add bounded event replay for forecast-catalog projections with checkpoints.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.67,"slice":"core","lang":"en"}
|
|
{"prompt":"Create the responsive workflow-template gallery with filters, versions, inputs, preview, and accessible selection.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.61,"slice":"core","lang":"en"}
|
|
{"prompt":"fix the map colors","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.34,"slice":"vague-eval","lang":"en"}
|
|
{"prompt":"Replace `GridEvt` with `GridEvent` across packages while preserving event names and schemas.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.56,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Diagnose why completed runs still trigger stalled-job alerts, implement the verified projection fix, and update runbook documentation.","purpose":"debugging","secondary":"writing","mixed":true,"difficulty":0.73,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Why does the catalog retain two previous metadata schemas? Explain current compatibility and rollback behavior.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.47,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Track down occasional missing chunks in Zarr dataset reads.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.69,"slice":"core","lang":"en"}
|
|
{"prompt":"Write a troubleshooting guide for failed workflows, covering staging, quotas, environments, checkpoints, and publication.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.43,"slice":"core","lang":"en"}
|
|
{"prompt":"Security review packet:\n\n```text\nFeature: shareable research result link\nCreator selects one published result and expiry up to 30 days\nRecipient does not need an account\nLink grants read-only access to plots, metadata, and selected output files\nEmbargoed datasets are excluded by UI but not independently checked by link service\nToken appears in URL path\nCreator may revoke early\nDownloads are logged with truncated IP and user agent\nLinked files may reveal paths containing project names\nCDN caches successful responses for one hour\n```\n\nReview embargo enforcement, token leakage, file scope, path metadata, revocation, caching, privacy, and audit. No implementation.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.77,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Add a loading state to the run cards","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.23,"slice":"core","lang":"en"}
|
|
{"prompt":"Plan and implement project-defined data retention. First define precedence, published snapshots, citations, holds, scratch data, deletion proof, and rollout; then add policy storage and enforcement.","purpose":"planning","secondary":"backendImpl","mixed":true,"difficulty":0.94,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Outline an architecture for federated scientific-data search across institutions, including identity, authorization, metadata, ranking, and phased adoption.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.91,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Implement snapshot-consistent experiment exports with resumable chunks.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.68,"slice":"core","lang":"en"}
|
|
{"prompt":"Escribe una guía para integrar observaciones científicas, con unidades, tiempos, calidad, correcciones y pruebas.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.4,"slice":"core","lang":"es"}
|
|
{"prompt":"Move dataset-license resolution out of the download handler without changing access.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.58,"slice":"core","lang":"en"}
|
|
{"prompt":"Investigate why the iOS field app loses queued samples after restoring a backup.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.63,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Add atomic variable-registry publication with validation, approval, and rollback.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.66,"slice":"core","lang":"en"}
|
|
{"prompt":"Remove the obsolete `legacy_grid` flag","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.04,"slice":"core","lang":"en"}
|
|
{"prompt":"Build the dataset-version comparison view with schemas, files, metadata, provenance, and keyboard navigation.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.66,"slice":"core","lang":"en"}
|
|
{"prompt":"`Temprature anomaly` を `Temperature anomaly` に直して。新しい英語ラベルだけです。","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.03,"slice":"boundary","lang":"ja"}
|
|
{"prompt":"Review the public raster-tile endpoint for embargo leakage, expensive ranges, crafted coordinates, cache confusion, and tenant bypass.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.75,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement model-run comparison history and write its API guide. Include runs, variables, metrics, artifacts, pagination, and permissions.","purpose":"backendImpl","secondary":"writing","mixed":true,"difficulty":0.7,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Plan a disaster-recovery strategy for scientific workflows, including datasets, environments, checkpoints, catalogs, keys, drills, and RTO tiers.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.94,"slice":"core","lang":"en"}
|
|
{"prompt":"Add content-hash deduplication to uploaded observation files with provider provenance.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.62,"slice":"core","lang":"en"}
|
|
{"prompt":"Create the desktop workflow editor with DAG layout, parameters, validation, versioning, and keyboard controls.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.71,"slice":"core","lang":"en"}
|
|
{"prompt":"continue the dataset cleanup","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.44,"slice":"vague-eval","lang":"en"}
|
|
{"prompt":"Rename `RunRev` to `RunRevision` across packages while preserving serialized fields.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.56,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Investigate why revoked collaborators still receive workflow completion emails, implement the verified recipient fix, and update user documentation.","purpose":"debugging","secondary":"writing","mixed":true,"difficulty":0.76,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Assess the proposed notebook-result cache for project isolation, mutable environments, stale datasets, credentials, and invalidation.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.72,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Find why a workflow worker retries a permanently unsupported GPU request forever.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.64,"slice":"core","lang":"en"}
|
|
{"prompt":"Draft a support guide for recovering an interrupted dataset publication.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.4,"slice":"core","lang":"en"}
|
|
{"prompt":"Projection failure output:\n\n```text\ndataset=d_881 revision=14 publication event offset=99218\ncatalog projection writes status=published and searchable=true\ncheckpoint transaction fails after database failover\nDOI worker observes published state and registers identifier\nprojection restarts from previous offset\nreplays publication against already-published state and rejects duplicate transition\npartition stops; DOI exists but publication event remains uncheckpointed\n```\n\nDetermine how projection, checkpoint, and DOI registration lack one durable boundary and why replay cannot recognize the prior application.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.82,"slice":"pasted-context","lang":"en"}
|
|
{"prompt":"Add source count below the dataset title","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.23,"slice":"core","lang":"en"}
|
|
{"prompt":"Restructure the dataset catalog and document its metadata-provider interface. Preserve results while separating ingest, normalize, validate, index, and publish.","purpose":"refactor","secondary":"writing","mixed":true,"difficulty":0.8,"slice":"mixed","lang":"en"}
|
|
{"prompt":"Architect a global forecast-delivery platform with model ownership, regional caches, publication, corrections, and migration milestones.","purpose":"planning","secondary":null,"mixed":false,"difficulty":0.95,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Implement signed dataset receipts with schema, file count, and content digests.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.68,"slice":"core","lang":"en"}
|
|
{"prompt":"Write an onboarding guide for adding a new quality-control rule, including inputs, flags, fixtures, review, and rollout.","purpose":"writing","secondary":null,"mixed":false,"difficulty":0.45,"slice":"core","lang":"en"}
|
|
{"prompt":"Consolidate variable-name formatting across portal and exports without changing output.","purpose":"refactor","secondary":null,"mixed":false,"difficulty":0.54,"slice":"core","lang":"en"}
|
|
{"prompt":"Investigate why corrected station coordinates remain stale in map search.","purpose":"debugging","secondary":null,"mixed":false,"difficulty":0.61,"slice":"boundary","lang":"en"}
|
|
{"prompt":"Add transactional project transfer with dataset permissions and immutable provenance.","purpose":"backendImpl","secondary":null,"mixed":false,"difficulty":0.71,"slice":"core","lang":"en"}
|
|
{"prompt":"Pin NumPy to `2.3.2`","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.06,"slice":"core","lang":"en"}
|
|
{"prompt":"Build the mobile experiment-status screen with stages, resources, logs, outputs, cancellation, and accessible summaries.","purpose":"frontendImpl","secondary":null,"mixed":false,"difficulty":0.62,"slice":"core","lang":"en"}
|
|
{"prompt":"把 `Forcast run` 改成 `Forecast run`，不要修改 key。","purpose":"quickFix","secondary":null,"mixed":false,"difficulty":0.03,"slice":"boundary","lang":"zh"}
|
|
{"prompt":"Inspect the dataset-archive importer for path traversal, decompression bombs, schema spoofing, huge metadata, and tenant confusion.","purpose":"review","secondary":null,"mixed":false,"difficulty":0.76,"slice":"core","lang":"en"}
|
|
{"prompt":"Implement workflow-template history and write its API documentation. Include versions, validation, publication, rollback, pagination, and permissions.","purpose":"backendImpl","secondary":"writing","mixed":true,"difficulty":0.7,"slice":"mixed","lang":"en"}
|