Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Brycewang Stanford/Compile Latex — measured AI agent skill | DecimalAI

brycewang-stanford/compile-latex

SkillScore86Maintained

Use when compiling a LaTeX paper, debugging LaTeX errors, building a paper PDF, or when a .tex file fails to produce output. Handles engine detection (xelatex vs pdflatex), bibliography systems (biber vs bibtex), and multi-pass compilation.

by brycewang-stanford

Links this skill to your agents and keeps it current with the author’s updates.

On-demandOn-demand: invoked explicitly — never occupies your agent's context until called.

Doc Ops pack

decimalai skills pull brycewang-stanford-compile-latex

Get this skill — curl · raw file

curl -fsSL https://app.decimal.ai/s/brycewang-stanford-compile-latex/SKILL.md

Raw file: /s/brycewang-stanford-compile-latex/SKILL.md ↗ · machine-readable summary at /s/brycewang-stanford-compile-latex.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 25h ago

—-23% tokens0% turns86% of whole cases pass with skill (19/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-15	✗→✓	▲ Improved	-67%	0%
case-04	✓→✓	＝ Same ✓	24%	0%
case-05	✓→✓	＝ Same ✓	11%	0%
case-06	✓→✓	＝ Same ✓	21%	0%
case-07	✓→✓	＝ Same ✓	-55%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/2/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

brycewang-stanford / Auto-Empirical-Research-Skills

skills/60-regisely-superpapers/skills/compile-latex/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/2/2026Version v3Bundled 1 file · scripts

Skill contents

SKILL.md

Compile LaTeX Manuscript

Compile a LaTeX manuscript using the standard 3-pass sequence for full citation resolution.

Steps

Identify the file: Use $ARGUMENTS as the base name (without .tex). If no argument, look for the main .tex file in the paper folder.

Compile (from the paper folder):

bash pdflatex -interaction=nonstopmode $ARGUMENTS.tex bibtex $ARGUMENTS pdflatex -interaction=nonstopmode $ARGUMENTS.tex pdflatex -interaction=nonstopmode $ARGUMENTS.tex

Check results:

Grep for Overfull \\hbox warnings -- count them
Grep for undefined citations or missing references
Verify PDF was generated with non-zero size: ls -la $ARGUMENTS.pdf
Report page count if pdfinfo is available

Report:

Compilation: PASS / FAIL
Warnings: N overfull hbox, N undefined citations
PDF: exists, N pages, X KB

Notes

Uses pdflatex by default. Change to xelatex or lualatex if your project requires it.
The .bib file should be in the same directory or referenced with a relative path.
Parameter estimates loaded via \input{} must exist at their referenced paths.

Want to see if brycewang-stanford/compile-latex would activate on your phrasing? Try it before you install — no LLM call, no signup.

Other measured skills in the registry, with their headline benchmark lift.