Profile My skills Saved My plugins Edit profile Submit a skill

← all publishers

ViderStar

@viderstar source repo

7 published skills

  1. ▌
    Jira Tasks · viderstar bundle
    Create and manage Jira Cloud issues through the REST API - file a task, bug, epic or sub-task, attach it to an epic and a sprint, fill in custom fields by their human names, set the assignee, link issues, change status, search with JQL. Works with any Atlassian Cloud site and any project; the project is a flag or a config value. Use whenever asked to create, file, update, assign, link, comment on or search Jira issues or tickets. Not for Confluence pages - that is the confluence-kb skill.
    0 installs
  2. ▌
    Confluence Kb · viderstar bundle
    Read and edit a Confluence Cloud knowledge base through the REST API - pages, spaces, CQL search, comments, with Markdown in and Markdown out. Use whenever asked to open, read, fetch, export, search, create, edit, update, rename, append to or comment on a Confluence or wiki page, including when given a .atlassian.net/wiki/ link. Not for Jira issues - that is the jira-tasks skill.
    0 installs
  3. ▌
    Dag Review · viderstar
    Review an Airflow DAG before it is merged: idempotency, retries and timeouts, failure alerting, heavy imports at module level, configuration through Variables, secrets in Connections, and concurrency settings. Use when asked to "review this DAG", "check the pipeline", "is this ready to merge", "why does the DAG fail", "add a new DAG", "optimise this DAG". Not for research notebooks.
    0 installs
  4. ▌
    Ab Analysis · viderstar
    Design and read A/B tests: sample size and MDE before the start, validity checks on the split, reading the result with a correction for multiple comparisons, and writing the conclusion for a business audience. Use when asked to "compute the MDE", "how much traffic do we need", "analyse this A/B", "look at the test results", "is it significant", "design an experiment", "why did the groups diverge". Not for offline evaluation on historical data, and not for rolling an arm out to production.
    0 installs
  5. ▌
    Experiment Article · viderstar bundle
    Write up an ML experiment from the notebook and the run log, in a skeleton that keeps the failures: verdict in the title, protocol as a table, ranking and business metrics side by side, "What was not verified", and a Reproduction section. Use when asked to "write up the experiment", "document the results", "describe what came out of the notebook", "summarise the run", "prepare the report on this experiment". Not for READMEs and not for code documentation.
    0 installs
  6. ▌
    Ch Query · viderstar
    Query ClickHouse without getting an inflated number or melting a production node: deduplication on ReplacingMergeTree, partition pruning with PREWHERE, column pruning, join order, and how to explore an unfamiliar schema first. Use when asked to "count this from the data", "how many orders/sales/events", "pull that out of ClickHouse", "write a query", "what tables are there", "check this hypothesis against the data", "why is this query hanging", "optimise this query", and before any analysis that starts with "how many do we have". Not for designing tables and migrations, not for A/B test analysis.
    0 installs
  7. ▌
    Nb Run · viderstar bundle
    Run a Jupyter notebook headless on a remote machine through nbconvert with a named profile of env flags: pre-flight check, launch detached from ssh, notification when it finishes, failure triage and comparison against the previous run. Use when asked to "run this notebook", "run the experiment", "rerun with different parameters", "train on the server", "run the smoke/cheap variant", "watch the run", "tell me when it's done", "why did the notebook fail", "what's the status of the run", and when reproducing a run from an experiment write-up. Not for editing notebooks interactively and not for triggering Airflow DAGs.
    0 installs