SkillsBench is a benchmark for measuring how effectively AI agents use modular skills—folders containing instructions, scripts, and resources—to complete specialized tasks. It helps researchers and developers evaluate both skill quality and agent behavior, including tasks that require combining multiple skills. The catalogue’s skills and instructions are evaluated as part of this workflow.
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 benchflow-ai/skillsbench --skill academic-pdf-redactiongit clone --depth 1 https://github.com/benchflow-ai/skillsbenchWrote 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/benchflow-ai/skillsbench/academic-pdf-redaction)<a href="https://agentmods.dev/skills/benchflow-ai/skillsbench/academic-pdf-redaction"><img src="https://agentmods.dev/badge/skills/benchflow-ai/skillsbench/academic-pdf-redaction/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/benchflow-ai/skillsbench/academic-pdf-redaction"><img src="https://agentmods.dev/badge/skills/benchflow-ai/skillsbench/academic-pdf-redaction.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00017 | $0.00966 |
| Opus 5 | $0.00009 | $0.00483 |
| Sonnet 5 | $0.00003 | $0.00193 |
| Haiku 4.5 | $0.00002 | $0.00097 |
Grade A, and why
academic-pdf-redaction 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 7d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- academic-pdf-redaction — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PDF Redaction for Blind Review
Redact identifying information from academic papers for blind review.
CRITICAL RULES
- PRESERVE References section - Self-citations MUST remain intact
- ONLY redact specific text matches - Never redact entire pages/regions
- VERIFY output - Check that 80%+ of original text remains
Common Pitfalls to AVOID
# ❌ WRONG - This removes ALL text from the page:
for block in page.get_text("blocks"):
page.add_redact_annot(fitz.Rect(block[:4]))
# ❌ WRONG - Drawing rectangles over text:
page.draw_rect(fitz.Rect(0, 0, 600, 100), fill=(0,0,0))
# ✅ CORRECT - Only redact specific search matches:
for rect in page.search_for("John Smith"):
page.add_redact_annot(rect)
Patterns to Redact (Before References Only)
IMPORTANT: Use FULL names/phrases, not partial matches!
- ✅ "John Smith" (full name)
- ❌ "Smith" (partial - would incorrectly match "Smith et al." citations in References)
- Author names - FULL names only (e.g., "John Smith", not just "Smith")
- Affiliations - Universities, companies (e.g., "Duke University")
- Email addresses - Pattern:
*@*.edu,*@*.com - Venue names - Conference/workshop names (e.g., "ICML 2024", "ICML Workshop")
- arXiv identifiers - Pattern:
arXiv:XXXX.XXXXX - DOIs - Pattern:
10.XXXX/... - Acknowledgement names - Names in "Acknowledgements" section
- Equal contribution footnotes - e.g., "Equal contribution", "* Equal contribution"
PyMuPDF (fitz) - Recommended Approach
import fitz
import os
def redact_with_pymupdf(input_path: str, output_path: str, patterns: list[str]):
"""Redact specific patterns from PDF using PyMuPDF."""
doc = fitz.open(input_path)
original_len = sum(len(p.get_text()) for p in doc)
# Find References page - stop redacting there
references_page = None
for i, page in enumerate(doc):
if "references" in page.get_text().lower():
references_page = i
break
for page_num, page in enumerate(doc):
if references_page is not None and page_num >= references_page:
continue # Skip References section
for pattern in patterns:
# ONLY redact exact search matches
for rect in page.search_for(pattern):
page.add_redact_annot(rect, fill=(0, 0, 0))
page.apply_redactions()
os.makedirs(os.path.dirname(output_path), exist_ok=True)
doc.save(output_path)
doc.close()
# MUST verify after saving
verify_redaction(input_path, output_path)
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.
- 7d ago First seen · 113 lines · 17 tokens per session scan A 587dbb503749
academic-pdf-redaction is a skill published in the GitHub repository benchflow-ai/skillsbench (1,757 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 17 tokens to every session and 966 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 2026-09-03.
Other skills, from other repositories
document-organization-pipeline
End-to-end pipeline for extracting, classifying, and organizing documents by subject.
pdf-calendar-parsing
Extract calendar events, blocks, and time slots from PDF calendar files using pdfplumber.
file-organizer-by-subject
Organize files into subject folders using keyword-based classification of titles and abstracts, with fallback to full text extraction.
pdf-form-filling
Fill PDF form fields programmatically using Python libraries like pypdf or pdfrw.
pdf-text-extraction
Extract text content from PDF files for analysis and classification.
pdf-calendar-extractor
Extract text and identifying colored regions (e.g., rectangles) from a PDF using pdfplumber.