📦

pathml

Name: pathml
Author: davila7

Content revision r2 High Risk ⚡ Contains scripts⚙️ External commands📁 Filesystem access🌐 Network access

Analyze Pathology Images with PathML

Whole-slide workflows involve complex formats, preprocessing choices, and large datasets. This skill provides focused PathML guidance for reproducible pathology analysis.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "pathml" from https://skillstore.io/skills/davila7-pathml.md and its manifest at https://skillstore.io/api/skills/davila7-pathml/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/davila7-pathml.md Signed manifest GET /api/skills/davila7-pathml/manifest Signed lockfile GET /api/skills/davila7-pathml/lockfile

Test it

Using "pathml". Create a preprocessing plan for H&E whole-slide images.

Expected outcome:

Load each slide at a suitable pyramid level and generate overlapping tiles.
Detect tissue, normalize stain appearance, label artifacts, and preserve masks.
Store tiles, masks, metadata, and provenance in HDF5 for reproducible processing.

Using "pathml". Outline a CODEX segmentation workflow.

Expected outcome:

Collapse imaging runs and identify validated nuclear and membrane channels.
Segment cells locally with Mesmer, then quantify markers for each cell.
Export measurements and spatial coordinates to AnnData for downstream analysis.

Using "pathml". Recommend an evaluation plan for HoVer-Net.

Expected outcome:

Split data by patient before training to prevent information leakage.
Measure Dice, aggregated Jaccard, and panoptic quality on held-out slides.
Review errors by tissue type and confirm performance with pathology experts.

Security Audit

High Risk

v10 • 7/23/2026 Open versioned report

Most static detections are documentation syntax or standard local PathML operations, not executable threats. Remote DeepCell inference creates a real data-disclosure boundary, while unpinned installation and unrestricted checkpoint loading require safeguards.

Files scanned

4,143

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (2)

High

references/data_management.md:441

Unsafe Model Checkpoint Deserialization

The feature extraction example loads a PyTorch checkpoint with torch.load without restricting deserialization or establishing checkpoint trust.

torch.load can deserialize executable pickle content in affected configurations. The example provides no weights-only setting, provenance check, or hash verification.

Medium

SKILL.md:79-84

Unpinned Third-Party Package Installation

The quick start installs PathML and all optional dependencies without a version or integrity constraint, increasing dependency substitution and update risk.

The documented commands explicitly install unpinned packages from the configured package index. The behavior is legitimate, but dependency contents can change.

Capability review items (2)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium

references/multiparametric.md:175

Hardcoded URL

api_url='https://deepcell.org/api/predict',

SegmentMIFRemote sends pathology image channels to a fixed external DeepCell endpoint. The example does not warn about patient-data authorization or deidentification.

Medium

references/preprocessing.md:355

Hardcoded URL

api_url='https://deepcell.org/api'

The documented remote segmentation transform sends image data to DeepCell. Pathology images may be sensitive, and the example omits privacy and retention safeguards.

Risk Factors

⚡ Contains scripts (8)

references/data_management.md:442 references/machine_learning.md:229 references/machine_learning.md:499 references/machine_learning.md:541 references/machine_learning.md:473-477 references/preprocessing.md:409-416 references/preprocessing.md:443-448 references/preprocessing.md:474-483

⚙️ External commands (28)

📁 Filesystem access (23)

🌐 Network access (2)

references/multiparametric.md:175 references/preprocessing.md:355

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/davila7-pathml/audits/10?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/davila7-pathml/security.svg)](https://skillstore.io/skills/davila7-pathml?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/davila7-pathml?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/davila7-pathml/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/davila7-pathml.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

davila7. (2026). pathml security audit report (audit version 10) [Author version unspecified]. Skillstore. https://skillstore.io/skills/davila7-pathml/audits/10

BibTeX citation

@techreport{davila7-davila7-pathml-2026,
  author = {davila7},
  title = {pathml security audit report (audit version 10)},
  institution = {Skillstore},
  year = {2026},
  number = {10},
  url = {https://skillstore.io/skills/davila7-pathml/audits/10},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "pathml security audit report (audit version 10)"
version: "unspecified"
type: report
authors:
  - name: "davila7"
date-released: "2026-07-23"
url: "https://skillstore.io/skills/davila7-pathml/audits/10"
identifiers:
  - type: other
    value: "skillstore:davila7-pathml:audit:10"
    description: "Skillstore immutable audit report identifier"

Compare variants

2 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

K-Dense-AI Recommended

k-dense-ai-pathml

Skillstore Score 50

Evidence Confidence High

Skillstore usage 15

Updated

2026-09-09

davila7 Current

davila7-pathml

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-09-09

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Build a WSI preprocessing pipeline

Configure tiling, tissue detection, stain normalization, and HDF5 output for a pathology dataset.

Prepare spatial graph analysis

Convert segmented cells into feature-rich graphs for neighborhood analysis or graph neural networks.

Analyze multiplex imaging data

Plan CODEX or Vectra segmentation, marker quantification, AnnData export, and batch processing.

Try These Prompts

Explain PathML basics

Explain how PathML loads a {slide_format} image. Compare pyramid levels, tiles, and regions for a beginner.

Design a preprocessing pipeline

Design a PathML pipeline for {stain_type} slides using {tile_size}-pixel tiles. Include quality control, normalization, segmentation, and output storage.

Plan multiplex analysis

Plan a PathML workflow for {platform} data with {nuclear_channel} and {membrane_channel}. Include segmentation, marker quantification, AnnData export, and privacy safeguards.

Optimize model training

Design and evaluate a {model_name} workflow for {task}. Address data splits, augmentation, metrics, checkpoints, ONNX deployment, and compute constraints.

Best Practices

Inspect slide metadata and pyramid dimensions before selecting tile size, stride, and level.
Keep raw slides immutable, record preprocessing parameters, and validate outputs on representative tissue.
Deidentify clinical images and review data policies before using remote inference services.

Avoid

Do not process every slide at full resolution without estimating memory, storage, and runtime.
Do not evaluate pathology models with training data or omit patient-level data separation.
Do not send identifiable images to external APIs without authorization and documented safeguards.