iclr-submission — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited iclr-submission (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Use this for an ICLR main-conference submission audit. Reopen the current CFP, Author Guide, OpenReview submission form, template, Code of Ethics, Code of Conduct, and LLM policy before giving deadline-ready advice.
or later publication track.
current freeze/change windows.
rely on a placeholder.
used 9 pages of main text, references outside the limit, and appendices after references.
metadata, comments, acknowledgments, and file names.
explicit exemption.
undisclosed LLM-written scientific content can become ethics issues.
ICLR enforcement is strict and the consequences are public: even a withdrawn submission leaves a visible OpenReview trace. Triage the highest-probability rejections first.
| Risk | Quick test | Fix before deadline |
|---|---|---|
| De-anonymization | Grep PDF and ZIP for names, remotes, buckets | Strip metadata, neutralize links |
| Over-limit main text | Count pages against the current rule | Move material after references |
| Author-list change | Check the freeze window in the CFP | Finalize authors before the cutoff |
| Dual submission | Confirm no archival overlap | Withdraw or rescope the other venue |
| LLM disclosure | Confirm the form field is truthful | Disclose; remove hallucinated citations |
A team submits an LLM fine-tuning paper whose anonymized repo link points to a GitHub org naming the lab, with a caption thanking a named cluster. Both leak identity. The fix: replace the link with an anonymized supplement ZIP, remove the acknowledgment until camera-ready, re-grep the PDF for the cluster name, and confirm all authors completed OpenReview profiles before the freeze.
[ICLR readiness] Ready / Needs fixes / Not ready
[OpenReview state] profiles / abstract / paper / supplement / reviewer nomination
[Blocking risks] <anonymity/page-limit/dual-submission/LLM/reciprocal-reviewing>
[Fix order] <highest-impact fixes before deadline>
[Current-cycle pages to reopen] <CFP, Author Guide, OpenReview group, template, LLM policy>~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.