📦

esm

Name: esm
Author: davila7

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

Design and Analyze Proteins with ESM

Protein modeling workflows often require unfamiliar ESM APIs, model choices, and generation settings. This skill provides practical guidance for local ESM models and scalable Forge inference.

Supports: Claude Codex Code(CC)

📊 67 Adequate

Install with my Agent

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

Agent request

Review the Skillstore skill "esm" from https://skillstore.io/skills/davila7-esm.md and its manifest at https://skillstore.io/api/skills/davila7-esm/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

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-esm.md Signed manifest GET /api/skills/davila7-esm/manifest Signed lockfile GET /api/skills/davila7-esm/lockfile

Test it

Using "esm". Recommend an ESM C workflow for clustering 2,000 protein sequences.

Expected outcome:

Model: start with ESM C 300M for throughput, then compare a larger model on a representative subset.
Processing: group similar lengths, batch on the available device, and cache pooled embeddings.
Analysis: normalize embeddings, cluster them, inspect representatives, and save assignments with reproducible parameters.

Using "esm". Plan structure prediction for a single enzyme sequence.

Expected outcome:

Use ESM3 with the structure track, record model and sampling settings, export a PDB, and assess confidence before experimental use.

Using "esm". Outline a reliable Forge batch generation job.

Expected outcome:

Submit bounded asynchronous batches and apply exponential backoff for rate limits.
Checkpoint completed items, cache deterministic results, and record failed requests for review.
Review provider data terms and apply biosafety screening before downstream synthesis or testing.

Security Audit

Medium Risk

v9 • 7/23/2026 Open versioned report

Eighty-four of 85 static findings are false positives caused by documentation syntax, scientific terms, fixed output examples, or expected Forge API references. One shortened community link is a low-risk redirect concern, and the protein generation workflows create a separate medium-severity dual-use biosafety risk.

Files scanned

2,682

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (1)

Medium

SKILL.md:14-22 SKILL.md:133-155 SKILL.md:298-300

Dual-use protein design capability

The skill provides actionable novel-sequence, inverse-folding, and function-conditioned generation workflows. Its general responsible-use note does not define safeguards for harmful biological functions.

The cited sections explicitly enable novel and function-conditioned protein generation, while the only biosafety control is a broad responsible-use statement.

Capability review items (1)

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

Low

SKILL.md:295

Hardcoded URL

- **Community:** Slack community at https://bit.ly/3FKwcWd

The shortened bit.ly community URL conceals its final destination and can redirect users if the link ownership or target changes.

Risk Factors

⚡ Contains scripts (1)

references/esm-c-api.md:338

🌐 Network access (14)

references/esm-c-api.md:575 references/forge-api.md:186 references/forge-api.md:368 references/forge-api.md:38 references/forge-api.md:80 SKILL.md:48 SKILL.md:245 SKILL.md:289 SKILL.md:290 SKILL.md:291 SKILL.md:293 SKILL.md:294 SKILL.md:295 SKILL.md:300

📁 Filesystem access (13)

references/esm-c-api.md:462 references/esm3-api.md:55 references/esm3-api.md:191 references/forge-api.md:328 references/forge-api.md:463 references/forge-api.md:438 references/workflows.md:515 references/workflows.md:92 references/workflows.md:95 references/workflows.md:278 references/workflows.md:283 references/workflows.md:436 references/workflows.md:634

⚙️ External commands (41)

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-esm/audits/9?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

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

HTML badge

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

Embed card

<iframe src="https://skillstore.io/embed/skills/davila7-esm.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). esm security audit report (audit version 9) [Author version unspecified]. Skillstore. https://skillstore.io/skills/davila7-esm/audits/9

BibTeX citation

@techreport{davila7-davila7-esm-2026,
  author = {davila7},
  title = {esm security audit report (audit version 9)},
  institution = {Skillstore},
  year = {2026},
  number = {9},
  url = {https://skillstore.io/skills/davila7-esm/audits/9},
  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: "esm security audit report (audit version 9)"
version: "unspecified"
type: report
authors:
  - name: "davila7"
date-released: "2026-07-23"
url: "https://skillstore.io/skills/davila7-esm/audits/9"
identifiers:
  - type: other
    value: "skillstore:davila7-esm:audit:9"
    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-esm

Skillstore Score 68

Evidence Confidence High

Skillstore usage 12

Updated

2026-08-21

davila7 Current

davila7-esm

Skillstore Score 67

Evidence Confidence High

Skillstore usage 8

Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: High

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Explore protein representations

Generate ESM C embeddings and plan similarity, clustering, or classification analyses for a sequence collection.

Prototype protein designs

Plan sequence completion, inverse folding, or function-conditioned ESM3 experiments with clear validation stages.

Scale inference pipelines

Design Forge workflows with asynchronous batches, retries, checkpoints, caching, and rate-limit controls.

Try These Prompts

Generate one embedding

Explain how to generate one ESM C embedding for [protein sequence]. Recommend a model, pooling strategy, device, and validation checks.

Predict a protein structure

Create a structure-prediction workflow for [sequence] using [local ESM3 or Forge]. Include model choice, generation settings, PDB output, and failure handling.

Plan inverse folding

Plan an inverse-folding experiment for [PDB file]. Explain sequence masking, sampling settings, candidate ranking, structural checks, and biosafety review.

Build a scalable Forge pipeline

Design a Forge pipeline for [dataset] and [task]. Include batching, retries, checkpointing, caching, cost controls, data handling, and biosafety review gates.

Best Practices

Prototype with smaller models, then compare larger models only when quality gains justify added cost.
Record model identifiers, generation parameters, seeds, input provenance, and validation results.
Screen generated sequences for biosafety concerns and require expert review before experimental validation.

Avoid

Do not treat generated sequences, structures, or function labels as experimentally validated results.
Do not place Forge tokens in source files, prompts, notebooks, or shared logs.
Do not send sensitive sequences to remote services without reviewing access controls and provider terms.