jinko-task-cmaes

CMA-ES Calibration Task

Execute a confirmed calibration specification. Do not invent objectives, constraints, parameter priors, optimizer options, or acceptance criteria.

PREREQUISITE: This skill needs an initialized jinko-sdk connection and an SDK satisfying its metadata.requires_sdk range. Run the jinko-sdk-setup skill (../jinko-sdk-setup/SKILL.md) and proceed only once its check passes. If that skill is not found, install it from novainsilico/jinko-skills.

Inputs

Require:

• a model SID;
• parameter priors with physical bounds;
• seed, thresholdWeightedScore, numberOfIterations, and populationSize;
• at least one fitness source: calibration-ready data tables and/or an advanced

output set containing objectives;

• any protocol, simple output set, advanced output set, folder, and name needed

by the specification.

If quantitative evidence has not yet been converted into a calibration-ready table, use jinko-task-extract-data-table. Use jinko-data-table, jinko-output-set, jinko-model, and jinko-protocol only for their respective Jinkō object mechanics.

Workflow

1. Resolve every input to its intended SID and snapshot. Present missing or

ambiguous inputs instead of guessing.

1. Use jinko-calibration-cmaes and its bundled creation script. Review its

dry-run output before applying it. The script owns parameter encoding, fitness-table eligibility, bound scaling, creation, and post-creation sanity; stop on an error and surface warnings.

1. Return the created calibration SID, revision, snapshot, URL, and effective

options for confirmation.

1. Use the lower-level run script to perform pre-launch sanity, launch, and wait

for a terminal state. Do not relaunch a terminal snapshot; create or update a configuration so the intended change has a new snapshot.

1. Use the lower-level inspection interfaces to collect the final status,

stopping reason, performance, results summary, objective weights, and the patient sorted first by optimizationWeightedScore when available. Fetch per-patient scalars, timeseries, errors, or augmented data tables only when their required selectors are present in the result metadata.

Return

Return:

• calibration SID, revision, snapshot, and URL;
• effective input references, priors, and optimizer options;
• sanity warnings, terminal status, stopping reason, and performance;
• supported result payloads and best-patient identity, with the iteration and

scenario arm needed for subsequent result calls;

• a concise account of unavailable requested outputs.

Do not claim a separate run ID, convergence analysis, score-evolution curve, best-patient parameter values, parameter posterior, or simulation-vs-data plot unless the returned API payloads actually provide the required data.