Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› ccs-review-process
ccs-review-process
GitHub指导ACM CCS论文审稿策略，涵盖双周期HotCRP流程、对抗性评审机制、伦理审查及轻微修订决策。提供评分杠杆表与分阶段应对建议，助力作者优化威胁模型、证据及伦理披露，提升录用概率。
Trigger Scenarios
咨询ACM CCS审稿流程
制定论文反驳或修改策略
理解CCS评审标准与伦理要求
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill ccs-review-process -g -y
SKILL.md
Frontmatter
{
"name": "ccs-review-process",
"description": "Use when explaining or planning around ACM CCS peer review, the two-cycle HotCRP workflow, decision categories including minor revision, the rebuttal window, the adversarial reviewer pool, ethics and disclosure scrutiny, meta-review dynamics, and ACM proceedings outcomes."
}
CCS Review Process
Use this to reason about review-stage strategy. Reopen the current CFP, the cycle's HotCRP site, the reviewer guidelines if posted, and the ethics policy before making process claims.
Process model
- CCS runs two review cycles per year on HotCRP, each with abstract registration, full-paper submission, review, a rebuttal window, and notification.
- Reviewers assess threat-model soundness, novelty, technical correctness, evidence quality, ethics and responsible disclosure, and significance to the security community.
- Decisions are not binary: CCS 2026 used accept, minor revision (authors revise for the same cycle under a shepherd or check), and reject. A first-cycle reject cannot return that year.
- An ethics or disclosure concern can sink an otherwise strong paper; the ethics reviewer's view carries weight independent of technical scores.
- Accepted papers publish in the ACM Digital Library proceedings, so camera-ready and metadata compliance matter as much as the initial decision.
Who reviews here
- The pool is adversarial by training: expect at least one reviewer to attack your threat model, look for the assumption that makes the attack easy, and probe whether the defense was tested against an adaptive attacker.
- CCS is broad, so a paper may draw reviewers from an adjacent sub-area; the threat model and intro must orient a web-security reviewer to a crypto result and vice versa.
- Borderline papers usually fall on one of three edges: an attacker assumption too strong to be interesting, a defense never adapted against, or an ethics gap left unaddressed.
Scoring leverage table
| Review dimension | What raises it | What sinks it |
|---|---|---|
| Threat model | A tight, realistic adversary stated up front | Capabilities that quietly grow to make the attack work |
| Novelty | An attack class or guarantee the literature lacked | An increment over a cited paper with no new idea |
| Evidence | End-to-end demonstration against a real target | Lab-only results against a strawman |
| Ethics | Disclosure done, harm minimized, reasoned in the paper | Real-world harm with no disclosure or IRB reasoning |
Stage-by-stage realism
- Initial reviews: triage by what the meta-reviewer will weigh — threat model and evidence — not by reviewer tone.
- Rebuttal: the window is short and structured; one crisp answer to the decision-critical objection beats an exhaustive point-by-point.
- Decision: the meta-review synthesizes; one unresolved threat-model or ethics objection outweighs several resolved clarity complaints.
- Minor revision: treat it as conditional acceptance with a checklist, not a debate; do exactly what the shepherd asks within the cycle's timeline.
Output format
[Current stage] submitted / reviews / rebuttal / decision / revision / camera-ready
[Decision actors] <reviewers / ethics reviewer / meta-reviewer / chairs>
[Likely leverage] <threat model / novelty / evidence / ethics>
[Forbidden moves] <identity leak / new unsupported results / ignoring the shepherd>
[Next response move] <one action>
Version History
- 9f86f09 Current 2026-07-19 14:16
