BRYANN2K/stipulate-skills / skills/stip-bootstrap
stip-bootstrap
VerifiedAdopt Stip in a new or existing repository, preserving project conventions and assessing applicable domains.
assets/extensions/data-engineering/check.md
View on GitHub# data-engineering — check
The points below are **candidates**, not a mandatory checklist or already approved criteria. During explore/validate, the agent selects and reformulates useful candidates as actual, unique `AC-n` criteria in `spec.md`. During check, verify only that contract; do not add requirements on the fly.
## Observable properties to adapt
- The specification identifies source, owner, grain, window/partition, output dataset, and consumers; unconfirmed information remains `incomplete`/`unverified` and cannot pass.
- Rerunning the same window produces expected output without duplicates or divergent effects, or an exclusion is agreed before approval for an explicitly non-replayable stream.
- Selected quality tests, such as schema, completeness, uniqueness, or freshness, run on a dated partition and their report is retained.
- The run and dataset are traceable with ownership, version, and required lineage; `failed`/`unverified` requires correction or revision and renewed approval.
During `validate`, the agent maps selected `AC-DATA-*` candidates to unique `AC-n` IDs in the combined specification; the engine does not perform this mapping.
## Reconcile evidence
For each `AC-n`, compare observations with the expected outcome, recording the command or protocol, environment/version, covered data, and limits. A file's existence, a test merely being written, or a tool finishing is not sufficient evidence of behavior. Distinguish simulated tests, real observations, and objectives requiring an operational observation period.
Counterexample to report: **The pipeline finishes without errors, but rows disappear during replay.** Link the discrepancy to the applicable criterion; correct it or report `failed` / `unverified`. A failure observation can be useful without satisfying a success criterion. Do not change criteria or thresholds to make them pass; contract changes require revision and renewed approval.
The core requires a report with the current subject digest and every exact criterion ID. The engine checks structure, statuses, and digests; it does not certify the truth of observations or domain relevance. Read the [evaluation cases](evaluation.json) to exercise selection and adoption judgment without presenting them as domain tests that have actually run.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |