infocom-workflow
GitHub用于规划IEEE INFOCOM论文提交全流程的项目管理技能。涵盖从选题、摘要注册、全文提交到录用通知及相机就绪的逆向时间规划，强调单夏季截止日期的严格日程安排与合规性检查。
触发场景
安装
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill infocom-workflow -g -y
SKILL.md
Frontmatter
{
"name": "infocom-workflow",
"description": "Use when planning an IEEE INFOCOM project timeline from venue fit through the two-step EDAS abstract-then-paper registration, the early-reject checkpoint, the December notification, the IEEE Xplore camera-ready with PDF eXpress and eCF, and presentation, with backward-planning offsets for a networking paper and honest handling of the single-summer deadline."
}
INFOCOM Workflow
Use this as the project-management skill for an INFOCOM submission. Replace every date with the current official timetable and work backwards from the abstract/metadata deadline (in late July), which precedes the full-paper deadline by about a week. INFOCOM runs a single annual summer deadline, so a missed cycle costs roughly a year — plan the calendar as tightly as the paper.
INFOCOM is an IEEE ComSoc conference with no standing editor-in-chief and no article-processing charge. Rotating leadership is the per-edition General Co-Chairs and TPC Co-Chairs, appointed under ComSoc; the cost model is conference registration, and at least one author must register for the paper to enter IEEE Xplore. Chairs rotate yearly; re-check the organizing-committee page rather than carrying a name forward.
Milestones
- Venue fit: confirm INFOCOM over SIGCOMM/NSDI/MobiCom and confirm the paper is ready this
cycle (
infocom-topic-selection). - Evidence lock: freeze the system model, theorems/proofs, simulator setup, baselines, and the evaluation results.
- Abstract/metadata registration: enter real title, abstract (≤200 words), co-authors, and topic tags into EDAS by the abstract deadline (INFOCOM 2026: 24 Jul 2025; 2027: 24 Jul 2026).
- Full-paper submission: upload the anonymized IEEEtran PDF (INFOCOM 2026: 31 Jul 2025; 2027: 31 Jul 2026), within the 10-page / ≤9-page-of-text budget.
- Early-reject checkpoint (if the cycle runs one): INFOCOM 2027 posts early-reject on 9 Oct 2026 — a mid-cycle signal, not a final verdict.
- Notification: final decision (INFOCOM 2026: 8 Dec 2025; 2027: 8 Dec 2026).
- Camera-ready: eCF copyright, PDF eXpress validation, IEEE Xplore-compliant PDF (INFOCOM 2026: 9 Jan 2026), plus author registration.
- Presentation: at least one author registers and presents at the conference.
Backward plan from the full-paper deadline
| Weeks out (heuristic) | Networking-paper milestone |
|---|---|
| 12+ | Problem formulated; system model and assumptions fixed; baselines chosen |
| 10 | Theorems proved / protocol designed; simulator or testbed built |
| 8 | Experiments/simulations complete; seeds and configs logged |
| 6 | Analysis done; plots and tables built; proofs written up |
| 4 | Full draft in IEEEtran; measured against the 9-page text boundary (refs excluded) |
| 3 | Internal mock review by a reader from the other side (theory ↔ systems) |
| 2 | Baselines hardened, related-work delta sharpened, page budget met |
| 1 | Double-blind sweep on the PDF and metadata; abstract entered in EDAS |
| 0 | Full paper on EDAS within the budget; five-paper cap re-checked |
These offsets are planning heuristics only — anchor every one to the current Call for Papers, never to a previous cycle's calendar.
The single-summer calendar reality
INFOCOM runs one annual research deadline in late July, with notification in December and the conference the following May. As of 2026-07-09, INFOCOM 2026 (Tokyo) is decided, so the live next target is INFOCOM 2027 (Honolulu; abstract 24 Jul 2026, paper 31 Jul 2026). Two planning consequences:
- A near-miss on the July deadline costs a full year at INFOCOM; a sibling flagship (SIGCOMM/NSDI/MobiCom) or a journal (ToN/JSAC) may have a nearer date — factor that into routing rather than idling.
- There is no rebuttal to plan for (traditionally): budget the pre-submission mock review as the one and only chance to hear "your baseline is unfair" from a friend instead of a reviewer.
Failure modes by stage
- Evidence still moving at week 4 forces last-minute simulation nobody has audited — the classic soundness reject.
- Discovering the paper is a page over at week 1 with appendices counted inside the nine pages — a not-reviewed outcome that editing weeks earlier would have prevented.
- Skipping the cross-side mock review forfeits the chance to catch an unstated assumption (theory reader) or an unrealistic topology (systems reader).
- Forgetting the five-paper cap on a group's submission sprint, so a sixth paper is dropped.
Coordination notes
- Assign one owner for the EDAS metadata/topic tags (they drive automated assignment) and another for the double-blind sweep; shared ownership is how both slip.
- Archive the exact submitted PDF and the simulator/testbed configuration — the camera-ready must match, and any released artifact must reproduce it.
Output format
[Current stage] idea / evidence / writing / abstract-registered / submitted / early-reject / accepted
[Next official deadline] <date and source, or unknown>
[Critical path] <three tasks that determine readiness>
[Risk register] <page budget / double-blind / evidence / baselines / author cap / registration>
[Owner map] <task -> person or role>
版本历史
- 9f86f09 当前 2026-07-19 16:02
