CtrlK
BlogDocsLog inGet started
Tessl Logo

paper-latex-sanitizer

Deterministically normalize safe LaTeX punctuation and replace unsupported forecast magnitudes with explicit placeholders before meta-paper-write publication gates run.

60

Quality

71%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./src/opensquilla/skills/bundled/paper-latex-sanitizer/SKILL.md
SKILL.md
Quality
Evals
Security

Paper LaTeX sanitizer

Internal deterministic pre-publication sanitizer for meta-paper-write. It runs before the length, citation, and publication-quality gates.

The sanitizer performs only two bounded repairs:

  • normalize U+2011, U+2013, and isolated U+2014 punctuation to LaTeX-safe ASCII equivalents while preserving paired Chinese em dashes (——);
  • when EVIDENCE_STATUS: not_supplied, replace concrete forecast magnitudes in results-facing sections with TBD / 待实验确定. Planned setup values remain intact. A declared result threshold remains intact only when the same magnitude is present in the authoritative user request/paper contract; labeling an invented number as a target or hypothesis does not preserve it;
  • remove only the known generated figure_placeholder_template or table_placeholder_template input line when the same manuscript already contains the corresponding inlined figure/table environment. Arbitrary user \input{...} commands and non-redundant template inputs remain untouched.

For an artifact-backed package, the referenced MANUSCRIPT_PATH is replaced atomically and the compact manifest is returned. Inline manuscript packages are normalized in the returned package. This step is not a quality gate: the independent paper-quality-gate must still run afterwards and block any unsupported semantic claims that remain.

Repository
TokenRhythm/opensquilla
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.