vldb-reproducibility
Use when engineering reproducibility into a VLDB paper before submission, covering hardware and configuration disclosure, dataset and workload provenance, run-to-run variance in systems measurements, competitor-version pinning, figure-to-raw-data traceability, and the disclosure floor PVLDB reviewers apply to performance claims.
Profile →npx skills add brycewang-stanford/Awesome-Journal-Skills --skill vldb-reproducibility --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# VLDB Reproducibility Use this while experiments are still running — reproducibility at a systems venue is an experimental-design property, not a packaging step. The question a PVLDB reviewer silently asks of every performance figure: *could a competent lab, given this paper alone, land within noise of these curves?* ## The disclosure floor Every performance claim needs its context recoverable from the paper (or its cited artifact): - **Hardware**: CPU model and count, memory, storage class and interface, network fabric, node count. "A commodity server" reproduces nothing. - **Software**: OS, kernel where it matters (I/O experiments), compiler and flags, and the exact versions *and configurations* of every system measured — including yours. - **Data**: source, size, skew characteristics; for generated data, the generator, its parameters, and its seed. - **Workload**: query mix, arrival pattern, client counts, warm-up protocol, and run duration. - **Measurement**: what was timed, from where, and what was excluded. ## Variance is a systems problem, not a seed problem ML papers randomize over seeds; systems papers fight nondeterminism from caches, compaction timing, JIT warm-up, ther
- The disclosure floor
- Variance is a systems problem, not a seed problem
- Competitor fairness ledger
- Traceability from figure to raw data
- The repro-honesty paragraph
- Output format
What does the vldb-reproducibility skill do?
Use when engineering reproducibility into a VLDB paper before submission, covering hardware and configuration disclosure, dataset and workload provenance, run-to-run variance in systems measurements, competitor-version pinning, figure-to-raw-data traceability, and the disclosure floor PVLDB reviewers apply to performance claims.
How do I install it?
Run `npx skills add brycewang-stanford/Awesome-Journal-Skills --skill vldb-reproducibility --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From brycewang-stanford/Awesome-Journal-Skills, a repository with 909 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.