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.
npx skills add sixb-ai/sixb --skill documentsgit clone --depth 1 https://github.com/sixb-ai/sixbWrote 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/skills/sixb-ai/sixb/documents)<a href="https://agentmods.dev/skills/sixb-ai/sixb/documents"><img src="https://agentmods.dev/badge/skills/sixb-ai/sixb/documents/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sixb-ai/sixb/documents"><img src="https://agentmods.dev/badge/skills/sixb-ai/sixb/documents.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00024 | $0.00664 |
| Opus 5.5 | $0.00010 | $0.00266 |
| Sonnet 5.5 | $0.00005 | $0.00133 |
| Haiku 4.5 | $0.00002 | $0.00066 |
Grade B, and why
documents scanned grade B with 1 finding 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 18d 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.
Asks the agent to reveal its instructionsmediumSystem prompt leakage
Directions to print, repeat or translate the system prompt extract configuration the operator did not intend to expose.
turn may need to edit the document. Publish only the requested formats using the worker's output instructions; preview PNGs and validation logs are working files unless the user requests them. What it actually says
The Sixb agent image has ReportLab, pypdf, pdfplumber, python-pptx, python-docx, openpyxl, pandas, PyArrow, NumPy, matplotlib, Pillow, PptxGenJS, LibreOffice, Poppler, Pandoc, qpdf, and DejaVu/Liberation/Noto fonts.
Use python3 for Python scripts. For JavaScript slide authoring, use a .cjs script with
const pptxgen = require("/opt/sixb/js/node_modules/pptxgenjs"); this works from any workspace.
Use the user's source facts, formatting, and template. Keep requested charts and tables editable.
Preserve distinctions such as proposed versus approved costs and response versus completion times.
For digital PDF tables, use pdfplumber and verify extracted rows and totals against the source. For scans, use pdf2image or pdftoppm followed by pytesseract/Tesseract; English OCR data is installed. OCR output can misread numbers and punctuation. Check source page images before using uncertain values. pypdf and pdfplumber alone do not OCR image-only pages.
Pandoc can read DOCX as Markdown or create DOCX from Markdown. For PDF output, render the generated DOCX with sixb-render; this image does not include a TeX engine. qpdf --check checks PDF structure, not visual layout. Use pandas/PyArrow for Parquet data; keep source units and null values intact.
After authoring, run:
sixb-render draft.pptx --output-dir previews-1
The helper accepts PDF, PPTX, DOCX, and XLSX. It creates a PDF, page PNGs, and a JSON manifest.
Always use a new output directory after revisions. Rendering has a 30-page default limit; use
--first-page and --last-page for larger documents and inspect every required page across batches.
The manifest's complete field says whether previews cover the entire document. Pass --timeout
(seconds per conversion step, default 60) when a large document times out.
Inspect the PNG paths with view_file when image input is supported. Check clipping, overlap,
contrast, fonts, chart labels, and units. Repair and render again when needed. Text extraction alone
does not verify layout. A model without image input can check text/structure but must not claim
visual inspection. LibreOffice is a compatibility check; it does not prove exact PowerPoint layout.
For spreadsheets, inspect formulas and values separately from printed-page previews. openpyxl does not calculate formulas. Use LibreOffice on a copy when recalculation is needed and verify the result.
Create previews inside the workspace so view_file can read them, and delete a preview directory
once you have inspected it: the workspace can outlive the turn. Keep source scripts when another
turn may need to edit the document. Publish only the requested formats using the worker's output
instructions; preview PNGs and validation logs are working files unless the user requests them.
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.
- 18d ago First seen · 48 lines · 24 tokens per session scan B 3108be18f8ac
documents is a skill published in the GitHub repository sixb-ai/sixb (115 stars, last pushed yesterday), licensed MIT. It adds 24 tokens to every session and 664 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it B with 1 finding (asks the agent to reveal its instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-28.
Other skills, from other repositories
figures
Makes publication-quality plots from data with matplotlib, learning curves, scaling laws, benchmark and ablation comparisons, Pareto trade-offs, heatmaps and confusion matrices, sized for the page, vector, with uncertainty shown. Use whenever results are plotted, charted or visualized for a paper, report or answer, or…
liteparse
Use this skill when the user asks to parse, perform multi-format document conversion or spatially extract text from an unstructured file (PDF, DOCX, PPTX, XLSX, images, etc.) locally without cloud dependencies.
Workspace Data Analyst
Analyze CSV files in the workspace and summarize insights.
smart-data-collection
A workflow for extracting structured information from images and documents such as PDFs, Word files, and spreadsheets, then storing it in a database.
meta-multi-format-export-pack
From one piece of source content, render four deliverables: .docx report, .pptx slides, .xlsx data, and an HTML/PDF public version.
superlinked-docs
Offload document, image, and structured-output work to the Superlinked inference cluster: convert PDF/DOCX/PPTX/XLSX/HTML/scans to clean markdown, describe an image (caption + tags), or produce schema/grammar-constrained JSON off the cluster — instead of ingesting the file directly, which can reduce the tokens billed…