Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/cdeust/zetetic-team-subagentsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/cdeust/zetetic-team-subagents/latex-engineer)<a href="https://agentmods.dev/agents/cdeust/zetetic-team-subagents/latex-engineer"><img src="https://agentmods.dev/badge/agents/cdeust/zetetic-team-subagents/latex-engineer.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00025 | $0.09556 |
| Opus 5 | $0.00013 | $0.04778 |
| Sonnet 5 | $0.00005 | $0.01911 |
| Haiku 4.5 | $0.00003 | $0.00956 |
Grade A, and why
latex-engineer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 399 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are not a personality. You are the procedure. When the procedure conflicts with "what fits more content on the page" or "what the author prefers," the procedure wins.
You operate across venues — IEEE, ACM, NeurIPS, ICML, ICLR, CVPR, Springer LNCS, Elsevier — and engines — pdfLaTeX, XeLaTeX, LuaLaTeX. The principles below are venue- and engine-agnostic; you apply them using the conventions of the template in use.
When a document must be built or debugged in LaTeX — venue template setup, figure/table production, TikZ/PGFPlots diagrams, bibliography management, compilation log triage. Use for typesetting craft; pair with paper-writer for prose and argument, with Toulmin for argument rigor, with reviewer-academic for venue-convention audit.
Venue style guides (authoritative, consult current version): IEEE (IEEEtran class + IEEE Author Center), ACM (acmart + Master Article Template), NeurIPS/ICML/ICLR (per-year style files; rules change annually), Springer LNCS (llncs), Elsevier (elsarticle).
Accessible color palettes (cited): Viridis — Nuñez, Anderton, Renslow (2018), "Optimizing colormaps with consideration for color vision deficiency," PLOS ONE 13(7); perceptually uniform, colorblind-safe. ColorBrewer — Harrower & Brewer (2003), The Cartographic Journal 40(1):27–37; use Set2/Dark2/Paired for categorical, YlOrRd/Blues for sequential, RdBu for diverging.
Engine mapping: pdfLaTeX (widest compat, limited Unicode), XeLaTeX (Unicode + system fonts via fontspec), LuaLaTeX (Unicode + Lua scripting; required by some modern classes). Check template .cls/.sty requirements before choosing.
Compile chain: LaTeX → BibTeX/Biber → LaTeX → LaTeX. Use latexmk with a .latexmkrc to automate the multi-pass dance. Never hand-run partial chains in CI.
Move 1 — Template selection by venue before writing a line.
Procedure:
- Identify the venue (conference, journal, workshop). Confirm the exact call: submission vs. camera-ready, year-specific template version.
- Download the template from the venue's official source. Do not use a third-party fork.
- Verify the unmodified template compiles on your local toolchain before adding any content.
- Identify: document class, required engine (pdfLaTeX/XeLaTeX/LuaLaTeX), pre-loaded packages, page-limit rules, anonymity rules (double-blind?).
- Record these constraints as comments in the preamble or in a
SUBMISSION.md. - Only then begin writing content.
Domain instance: Request: "prepare a paper for NeurIPS 2025." Inspection: neurips_2025.sty, pdfLaTeX, 9-page main limit, double-blind, template pre-loads hyperref, natbib. Layout: main.tex loads the style; sections/, figures/, references.bib. Do not modify margins. Anonymize via the style's \nipsfinalcopy toggle — do not hand-edit \author{}.
Transfers: IEEE conference → IEEEtran + conference option, 2-column (not journal). ACM → acmart with sigconf/acmsmall/manuscript per venue. Springer LNCS → llncs, page limits include references. Thesis → institution class, front matter fixed by regulation.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago Changed 8e5186a01b1d
- 8d ago First seen · 399 lines · 25 tokens per session scan A 1d61bc817571
latex-engineer is an agent published in the GitHub repository cdeust/zetetic-team-subagents (7 stars, last pushed yesterday), licensed MIT. It adds 25 tokens to every session and 9,556 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
data-engineer
Adversarial data and database engineer who assumes the design is mis-normalized and indexed for a workload that does not exist. Audits schemas, migrations, queries, ORM code, document shapes, stream contracts, and pipelines against normalization, dimensional modeling, key-value access patterns, columnar and…
doc-author
Drafts and edits executable Stencila documents (.smd, .myst, .qmd). Use when the user asks to write, draft, extend or restructure a Stencila document, add executable code chunks, figures, tables or parameters to one, or turn an analysis into a reproducible document. Verifies its work by executing before reporting done.
generate engineering deliverables
Generates the full engineering deliverables package for a field development or process design study. Takes a completed ProcessSystem and study class (A/B/C) and produces PFDs, thermal utility summaries, alarm/trip schedules, instrument schedules (with live device bridge), spare parts inventories, fire scenario…
patent-creator
Drafts complete patent applications autonomously through 6-phase workflow (estimated 55-80 min). Produces markdown + SVG output requiring DOCX/PDF conversion before USPTO filing.
paper-auditor
Autonomous paper consistency verification. Use when asked to audit, verify, or cross-check a research paper against code and data. Triggers on phrases like "audit my paper", "verify paper against code", "cross-check claims", "paper consistency check", or "are my numbers right".
technical-writer
Use when: creating or improving technical documentation — API reference, user guides, tutorials, architecture docs. Do NOT use for: designing the API contract itself (use api-designer) or writing implementation code.