Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› itcs-submission
itcs-submission
GitHub用于在上传至ITCS HotCRP前审计论文提交物。检查双盲匿名、格式规范（单栏11pt以上）、前10页核心内容展示及完整证明，确保符合开放获取与轻量级审稿要求。
Trigger Scenarios
准备提交ITCS会议论文
检查论文双盲匿名合规性
验证ITCS投稿格式与截止日期
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill itcs-submission -g -y
SKILL.md
Frontmatter
{
"name": "itcs-submission",
"description": "Use when auditing an ITCS submission for HotCRP readiness — the early-September abstract-then-paper deadlines, single-column >= 11pt format with no hard page limit, the first-10-pages merits window, complete proofs of all central claims, the lightweight double-blind sweep, and the archival-prior-publication check."
}
ITCS Submission
Run this audit before uploading to the ITCS HotCRP site. ITCS papers publish open access in
LIPIcs and are reviewed lightweight double-blind with no rebuttal, so the submitted PDF has
to be complete, self-defending, and correctly anonymized on the first and only try. Every rule
below was read from the ITCS 2026/2025 calls on 2026-07-09 via search renderings of the
itcs2026.hotcrp.com and itcs-conf.org URLs (see
../../resources/official-source-map.md); treat them as
a one-cycle snapshot and reopen the live call first.
The two-step deadline
ITCS separates the abstract/registration from the full-paper submission, and the abstract cutoff is earlier:
- Abstract / registration locks title, abstract, authors, and conflicts (ITCS 2026: 4 September 2025, 7:59:59 PM EDT). The abstract drives PC bidding — register the real title and abstract, not a placeholder.
- Full-paper submission uploads the anonymized PDF with complete proofs (ITCS 2026: 6 September 2025, 7:59:59 PM EDT). Miss the registration and no submission slot exists.
Format and page policy
- Single column, font size at least 11 point. Beyond that, the submission call imposes no
hard formatting template and no hard page limit (the LIPIcs
lipics-v2021class is a camera-ready requirement, not a submission one — seeitcs-camera-ready). - First-10-pages merits window. It is strongly recommended that the first ~10 pages give a
clear presentation of the paper's merits, significance, innovations, and place in the
literature. Reviewers judge the idea from this window; do not bury it (see
itcs-writing-style). - Complete proofs of all central claims must be included, in an appendix if needed. "Proof
omitted" or "see the full version" for a load-bearing claim is a substantive defect, not a
formatting one (see
itcs-supplementary).
Lightweight double-blind sweep
ITCS anonymity is lightweight: hide the authors in the PDF, but keep the references honest and allow a public full version. The leak surface is narrower than an empirical venue's (no code, no data), but real:
# Mechanical pass on the submission PDF
pdfinfo paper.pdf | grep -Ei 'author|creator|producer' # scrub metadata identity
pdftotext paper.pdf - | grep -nEi 'thank|acknowledg|grant|supported by|our (previous|prior|earlier) (work|paper)' | head
pdftotext paper.pdf - | grep -nEi '\\bwe (previously|earlier) (showed|proved)\\b' | head
- Remove author names, affiliations, emails, and acknowledgements/funding lines from the PDF.
- Neutralize first-person self-reference: rewrite "in our prior work [12]" as a third-person
"[Author, YEAR] showed." Do not anonymize or remove the reference itself — the reference
list stays intact (see
itcs-related-work). - A public arXiv/ECCC/ePrint full version is allowed and encouraged; you need not delay it for anonymity. Just keep the submitted PDF blinded.
Prior-publication and dual-submission check
- The result must not be published, presented, or under submission at another archival conference. Papers accepted to ITCS should not then go to another archival conference.
- Simultaneous submission to a journal is allowed — a distinctive ITCS latitude; a journal version in parallel is fine.
- Confirm the exact current wording each cycle; the archival/non-archival boundary and any workshop carve-outs are cycle-volatile.
Desk-risk triage
| Finding at audit time | Severity | Real fix |
|---|---|---|
| Central claim lacks a complete proof | Substantive reject-grade | Add the full proof (appendix); do not defer to the preprint |
| Merits not legible within ~10 pages | High | Restructure so model+motivation+results come first |
| Author identity in PDF metadata/acknowledgements | Anonymity violation | Scrub metadata; remove acknowledgements/funding |
| First-person self-citation ("our prior [12]") | Anonymity leak | Rewrite in third person; keep the reference |
| References anonymized/removed | Breaks the lightweight rule; reads evasive | Restore normal citations |
| Abstract/authors not registered by the earlier cutoff | No submission slot | Nothing fixes this post-deadline — calendar it now |
| Result already at another archival venue | Dual-submission violation | Withdraw or reroute; journal parallel is OK |
| Font < 11pt or multi-column | Formatting non-compliance | Recompile single-column, >= 11pt |
Final-week order of operations
- Freeze the model and definitions early; proofs can be polished, the framing cannot churn.
- Register title/abstract/authors/conflicts before the earlier registration cutoff.
- Run the proof-completeness pass: every central claim fully proved in the PDF.
- Run the mechanical anonymity checks on the final PDF (metadata + self-reference).
- Post the full version to arXiv/ECCC/ePrint if desired (allowed); keep the submission blinded.
- Fill every HotCRP field — topics, conflicts for every coauthor's institution and recent collaborators — a day early; late conflicts are the classic midnight failure.
- Re-download the uploaded PDF and read the first 10 pages cold to confirm the idea lands.
Reverify each cycle
- The two cutoff dates and their timezone (EDT/AoE).
- Whether the current call still runs no rebuttal and lightweight double-blind.
- The exact prior-publication/dual-submission wording and any generative-AI-disclosure rule.
- That there is still no hard page limit and the 11pt/single-column format holds.
Output format
[ITCS submission status] ready / blocked / needs work
[Registration] title/abstract/authors/conflicts locked by the earlier deadline? yes/no
[Format] single-column, >= 11pt? merits legible within ~10 pages? yes/no
[Proofs] every central claim fully proved IN the PDF (not only the preprint)? yes/no
[Anonymity] metadata clean, self-cites neutralized, references intact? yes/no
[Prior publication] not at another archival venue (journal parallel OK)? yes/no
[Fix queue] <ordered, with owners and dates before the early-September cutoff>
Version History
- 9f86f09 Current 2026-07-19 16:06
