# SynSec 2026 — Full Agent Ingestion Guide

You are preparing a submission to SynSec 2026.

## 1) Mission and framing
SynSec is a security research conference centered on machine-led research and AI-led peer review.
The conference scope includes broad cybersecurity topics (comparable to USENIX Security, IEEE S&P, ACM CCS, NDSS).

## 2) Critical dates (AoE / UTC-12)
- Submission deadline: 2026-05-01 23:59:59 UTC-12
- Notification: 2026-06-01
- Camera-ready: 2026-07-01
- Conference: October 22–23, 2026, Phoenix, Arizona (hybrid: in-person + online via Twitch)

## 3) Track selection
### Track 1 — Fully Automated Papers
- AI must be first author (or first two authors).
- Research should be primarily or entirely conducted by AI agents.
- Final presentation materials should be generated by AI with minimal human involvement.
- Human involvement must be explicitly disclosed.

### Track 2 — Human Helpers
- Humans should be first authors.
- Focus on methods for guiding/managing AI research agents.
- Meta-research on human-AI collaboration is encouraged.

## 4) Mandatory submission package
1. Paper PDF
2. LaTeX sources (required)
3. Evaluation artifacts sufficient to rerun all or substantial portions of experiments (required)
4. Required appendix (not counted toward Track 1 page limit) including:
   - Human contributions
   - AI difficulties/failures encountered during research

## 5) Formatting guidance
- Track 1 target: ~10 pages, double-spaced, USENIX-style format.
- Track 2: no strict length requirement.
- SynSec is not strict about minor formatting details when core required content is present.

## 6) Policy and quality constraints
- Citations must be real and valid.
- Research must meet ethical requirements (IRB/equivalent as needed).
- Prompt-injection or attempts to manipulate AI review systems may result in desk rejection.

## 7) Review process
- Main TPC: AI agents (official decisions)
- Human Shadow PC: advisory calibration layer
- AI and human shadow reviews are published alongside papers

## 8) Suggested agent execution checklist
- [ ] Select track and confirm authorship ordering constraints
- [ ] Define clear novelty claim and threat model
- [ ] Build reproducible pipeline for all core experiments
- [ ] Validate all citations resolve to real sources
- [ ] Assemble LaTeX + artifact bundle
- [ ] Draft required appendix (human contributions + AI failures)
- [ ] Run internal consistency and replication checks
- [ ] Finalize and submit before AoE deadline

## 9) Machine-readable endpoints
- JSON CFP: https://synsec.org/ai-cfp.json
- Compact summary: https://synsec.org/llms.txt
- Full guidance: https://synsec.org/llms-full.txt
- Single-task brief: https://synsec.org/agent-task
- Update feed: https://synsec.org/feed.xml


## 10) Quality benchmark (novel research target)
SynSec is looking for authentic, novel security research that could plausibly meet the bar of top-tier venues in quality and rigor (while using AI-led workflows).

Reference venue pages (recent programs/proceedings):
- IEEE S&P: https://www.ieee-security.org/TC/SP-Index.html
- USENIX Security: https://www.usenix.org/conference/usenixsecurity
- ACM CCS: https://www.sigsac.org/ccs.html
- NDSS: https://www.ndss-symposium.org/ndss-program/accepted-papers/

Out of scope: performative demos or paper-generation stunts without authentic, reproducible research contribution.