io-submission — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited io-submission (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
The last check before pressing submit on Editorial Manager. IO is double-blind, so the single most common avoidable failure is a manuscript that reveals the author through first-person self-references. Verify volatile specifics on the official page before relying on them.
(≤ 10,000). Word limits include tables, figures, and notes; exclude the bibliography.
word cap not numerically confirmed — 待核实).
work in the third person.
staff before final acceptance; Data Availability Statement before the reference list.
io-tables-figures)io-transparency-and-data-policy)Some failures are administrative, not intellectual, and IO's double-blind, word-counted process punishes them disproportionately. Rank your final pass by severity:
| Severity | Failure | Where it bites |
|---|---|---|
| Fatal | author identity leaks (first-person self-cite, acknowledgments, metadata) | double-blind integrity |
| Fatal | word count omits tables/figures/notes; paper is actually over cap | desk return for length |
| Medium | no Data Availability Statement before the references | transparency-policy gap |
The two fatal rows are the most common avoidable IO rejections; both are caught by a clean-room read of the anonymized PDF and a recomputed word count.
A treaty-compliance study is "ready." The recompute: body 12,400 words, but two dyadic tables and three figures add ~1,900 word-equivalents and footnotes another ~400, so the true count nears 14,700 — over the 14,000 Research Article cap (figures illustrative; recount on IO's stated basis). The fix is not prose-trimming alone: move one robustness table and the balance table into supplementary material (within the ~20-page norm). The clean-room pass then catches "as we showed in Author 2021" — a first-person self-reference — rewritten in the third person. Only then does it go to Editorial Manager. Confirm every numeric cap and the current portal against the journal's current submission guidelines before relying on them.
【Type】Research Article / Research Note / Essay (cap met? Y/N)
【Anonymized】third-person self-cites + metadata clean? [Y/N]
【Abstract/word-count/acknowledgments separate】[Y/N]
【Word count basis】incl. tables/figures/notes, excl. bibliography? [Y/N]
【Footnotes + ORCID + DAS】[Y/N]
【Repro + proofs staged for verification】[Y/N]
【Next】await decision → io-rebuttal on R&R../../resources/external_tools.md — reference managers, anonymization, repro tooling../../resources/official-source-map.md — official IO URLs behind every fact in this pack~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.