boltz

Structure prediction for protein, nucleic-acid, and small-molecule complexes with Boltz-2 (Passaro & Wohlwend et al. 2025, github.com/jwohlwend/boltz). Reach for this skill to validate designed binders against a target, to co-fold a protein with a SMILES or CCD ligand, or to get an open-source AlphaFold3 alternative with optional binding-affinity prediction.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

88%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, highly actionable body: complete input/output examples, concrete thresholds, and an error-recovery table with essentially zero padding. The only improvements are moving version-sensitive details out of the main flow and splitting deeper affinity/error material into reference files if the skill grows.

Suggestions

Move the 'Boltz v2.2.x caps affinity ligands at 128 atoms' version-pinned detail into a clearly labeled version-notes or compatibility section so it can be updated without touching the main flow.

If the skill expands, split the affinity-head and error-table detail into a one-level-deep reference file (e.g. references/affinity.md) with clear links from SKILL.md.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence (no explanations of what MSA, pTM, or diffusion are) with every section earning its place, but the inline version pin 'Boltz v2.2.x caps affinity ligands at 128 atoms' is time-sensitive detail not placed in a deprecation/old-patterns section, holding it just below the top anchor.	4 / 5
Actionability	Guidance is fully executable: a complete input YAML, a copy-paste 'boltz predict' command with flags, exact output paths, numeric interpretation thresholds, and an error-message-to-fix table covering the common failure cases.	5 / 5
Workflow Clarity	The flow is clearly sequenced — author YAML, run predict, read confidence_complex_model_0.json, apply explicit checkpoints (iptm > 0.5, complex_plddt > 0.7), then consult the error table for recovery — providing feedback loops without being a destructive or batch operation.	5 / 5
Progressive Disclosure	No bundle files exist and the body references none; sections are well-organized and self-contained at roughly 95 lines, but it exceeds the under-50-line simple-skill case and material like the affinity-head details and error catalogue could plausibly live in one-level-deep reference files.	4 / 5
	Total	18 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: concrete capabilities, an explicit 'Reach for this skill' trigger clause with three use cases, and a clearly delineated niche. The only gap is a few missing natural synonyms for the trigger vocabulary.

Suggestions

Add one or two common phrasing variants such as 'protein–ligand complex prediction' or 'binder design validation' to broaden natural trigger coverage.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete actions — 'validate designed binders against a target', 'co-fold a protein with a SMILES or CCD ligand', 'binding-affinity prediction' — covering the skill's capabilities comprehensively, matching the top anchor rather than the 'minor gaps' level 4.	5 / 5
Completeness	It explicitly answers both what ('Structure prediction for protein, nucleic-acid, and small-molecule complexes with Boltz-2') and when, with concrete trigger phrases ('Reach for this skill to validate designed binders... to co-fold a protein... or to get an open-source AlphaFold3 alternative').	5 / 5
Trigger Term Quality	Natural terms users would say are present ('binder', 'co-fold', 'SMILES', 'AlphaFold3', 'affinity'), but a few common variations like 'protein–ligand complex' or generic 'complex structure prediction' phrasings are missing, so it sits between the good (4) and comprehensive (5) anchors.	4 / 5
Distinctiveness Conflict Risk	It carves a clear niche (Boltz-2 co-folding and binder validation) with distinct trigger vocabulary (SMILES/CCD ligands, affinity prediction), making it unlikely to fire for unrelated skills.	5 / 5
	Total	19 / 20 Passed

Validation

75%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 missing	Warning

	Total	12 / 16 Passed

Repository: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/claude-science/boltz/SKILL.md
Commit: a16301c

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.