Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when deciding which computer-science or AI conference skill to invoke next, comparing fit across the 155-conference CS roadmap, or routing an AI/ML/CS manuscript before venue-specific re-framing.
.claude/skills/brycewang-stanford-cs-ai-conference-workflow/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 73% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 95% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 122% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 146% | 0% |
| case-22 | ✓→✗ | ▼ Worse | 112% | 0% |
This is the router for the computer-science conference pack. It puts AI conferences first, then routes by contribution type across ML, data mining, vision, NLP, robotics, HCI, systems, security, software engineering, programming languages, databases, and theory. It does not replace a single-conference profile; it selects the right profile and forces an official-cycle check before submission.
../../resources/worked-examples/venue-routing.md.../../resources/exemplars/selection-patterns.md.This file now carries the high-confusion sibling contrasts surfaced by clone-audit review (robotics, graphics/vision, NLP chapters, systems/networking, security, SE/PL/theory).
../../resources/conference-roster.md and../../resources/official-source-map.md to open the current official CFP, author kit, and submission policy for the chosen conference.
| Manuscript signature | Prefer skills | |---|---| | AI/ML first | neural-information-processing-systems / international-conference-on-machine-learning / international-conference-on-learning-representations / aaai-conference-on-artificial-intelligence / international-joint-conference-on-artificial-intelligence | | Data mining and web AI | acm-sigkdd-conference-on-knowledge-discovery-and-data-mining / the-web-conference / acm-international-conference-on-web-search-and-data-mining / acm-conference-on-recommender-systems | | Vision and multimodal media | computer-vision-and-pattern-recognition / international-conference-on-computer-vision / european-conference-on-computer-vision / acm-international-conference-on-multimedia | | NLP, speech, and IR | annual-meeting-of-the-association-for-computational-linguistics / conference-on-empirical-methods-in-natural-language-processing / interspeech / acm-sigir-conference-on-research-and-development-in-information-retrieval | | Robotics and embodied AI | ieee-international-conference-on-robotics-and-automation / ieee-rsj-international-conference-on-intelligent-robots-and-systems / robotics-science-and-systems / conference-on-robot-learning | | HCI and visualization | acm-chi-conference-on-human-factors-in-computing-systems / acm-symposium-on-user-interface-software-and-technology / acm-conference-on-computer-supported-cooperative-work-and-social-computing / ieee-visualization-conference | | Systems, networking, architecture, and HPC | acm-symposium-on-operating-systems-principles / usenix-symposium-on-operating-systems-design-and-implementation / acm-sigcomm / international-symposium-on-computer-architecture | | Security and privacy | ieee-symposium-on-security-and-privacy / usenix-security-symposium / acm-conference-on-computer-and-communications-security / network-and-distributed-system-security-symposium | | Software engineering, PL, and formal methods | international-conference-on-software-engineering / acm-international-conference-on-the-foundations-of-software-engineering / acm-sigplan-conference-on-programming-language-design-and-implementation / acm-sigplan-symposium-on-principles-of-programming-languages | | Databases and theory | acm-sigmod-international-conference-on-management-of-data / international-conference-on-very-large-data-bases / acm-symposium-on-theory-of-computing / ieee-symposium-on-foundations-of-computer-science |
| Confusable targets | Decision rule | |---|---| | NeurIPS vs ICML vs ICLR | Use NeurIPS for broad ML/AI reach, ICML for machine-learning method/theory discipline, and ICLR for representation/deep-learning/open-review fit. | | KDD vs ICDM vs SDM | KDD leans data-mining impact and applied discovery, ICDM broad IEEE data-mining methods, SDM mathematical/statistical data-mining rigor. | | CVPR vs ICCV vs ECCV | All require a vision contribution; current cycle, scope, and reviewer community decide, not acronym prestige alone. | | ACL/EMNLP vs NAACL/EACL | Separate core NLP method, empirical analysis, resource construction, and chapter-cycle fit before choosing. | | CHI vs UIST vs CSCW vs IUI vs VIS | User study, UI systems, social computing, intelligent-interface, and visualization claims need different evidence. | | S&P vs USENIX Security vs CCS vs NDSS | Pick by threat model, attack/defense evidence, ethics posture, systems/security community, and current CFP scope. | | ICSE/FSE vs ASE vs ISSTA vs SANER/ICSME | Broad SE, automation, testing/analysis, reengineering, and maintenance-history papers are not interchangeable. | | SIGMOD vs VLDB vs ICDE | Data-management systems, PVLDB-style database research, and IEEE data-engineering work have different submission mechanics; verify the current cycle. |
text[Top conference skill] <skill-name> [Alt 1] <skill-name> (reason) [Alt 2] <skill-name> (reason) [Do not submit to] <venue> (one-line mismatch reason) [Biggest current gap] novelty / evidence / proof / artifact / user study / ethics / format / official requirements [Next step] invoke <skill-name> for single-venue fit and current-cycle checks
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 19,097 | 15,411 | -19% | 1 | 1 | 0% | 2,343 | 4,052 | +73% | 0 | 0 | — |
case-02 | fail→pass | 16,342 | 13,910 | -15% | 1 | 1 | 0% | 1,931 | 3,758 | +95% | 0 | 0 | — |
case-03 | fail→pass | 15,191 | 17,348 | +14% | 1 | 1 | 0% | 1,932 | 4,298 | +122% | 0 | 0 | — |
case-04 | pass→pass | 14,735 | 15,794 | +7% | 1 | 1 | 0% | 2,289 | 3,698 | +62% | 0 | 0 | — |
case-05 | pass→pass | 12,606 | 16,631 | +32% | 1 | 1 | 0% | 1,999 | 3,961 | +98% | 0 | 0 | — |
case-06 | pass→pass | 17,024 | 19,156 | +13% | 1 | 1 | 0% | 2,816 | 3,897 | +38% | 0 | 0 | — |
case-07 | pass→pass | 16,075 | 23,791 | +48% | 1 | 1 | 0% | 2,262 | 4,851 | +114% | 0 | 0 | — |
case-08 | pass→pass | 12,775 | 7,832 | -39% | 1 | 1 | 0% | 2,015 | 3,340 | +66% | 0 | 0 | — |
case-09 | pass→pass | 17,239 | 14,873 | -14% | 1 | 1 | 0% | 2,070 | 3,826 | +85% | 0 | 0 | — |
case-10 | pass→pass | 13,652 | 14,281 | +5% | 1 | 1 | 0% | 1,976 | 3,521 | +78% | 0 | 0 | — |
case-11 | pass→pass | 12,925 | 12,137 | -6% | 1 | 1 | 0% | 2,098 | 3,330 | +59% | 0 | 0 | — |
case-12 | pass→pass | 12,239 | 7,174 | -41% | 1 | 1 | 0% | 1,994 | 3,290 | +65% | 0 | 0 | — |
case-13 | pass→pass | 18,443 | 14,742 | -20% | 1 | 1 | 0% | 2,192 | 3,672 | +68% | 0 | 0 | — |
case-14 | pass→pass | 10,837 | 11,937 | +10% | 1 | 1 | 0% | 1,770 | 3,215 | +82% | 0 | 0 | — |
case-15 | pass→pass | 20,315 | 13,798 | -32% | 1 | 1 | 0% | 2,357 | 3,600 | +53% | 0 | 0 | — |
case-16 | pass→pass | 9,864 | 8,974 | -9% | 1 | 1 | 0% | 1,647 | 3,191 | +94% | 0 | 0 | — |
case-17 | pass→pass | 21,175 | 13,911 | -34% | 1 | 1 | 0% | 2,465 | 4,264 | +73% | 0 | 0 | — |
case-18 | pass→pass | 16,128 | 15,522 | -4% | 1 | 1 | 0% | 1,638 | 4,263 | +160% | 0 | 0 | — |
case-19 | fail→fail | 14,610 | 10,720 | -27% | 1 | 1 | 0% | 2,323 | 3,232 | +39% | 0 | 0 | — |
case-20 | pass→fail | 14,394 | 17,411 | +21% | 1 | 1 | 0% | 2,097 | 5,165 | +146% | 0 | 0 | — |
case-21 | pass→pass | 16,286 | 19,270 | +18% | 1 | 1 | 0% | 2,844 | 4,351 | +53% | 0 | 0 | — |
case-22 | pass→fail | 20,671 | 24,619 | +19% | 1 | 1 | 0% | 2,459 | 5,204 | +112% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.