CtrlK
BlogDocsLog inGet started
Tessl Logo

boltz-protein-design

Design new protein binders with Boltz. Use when generating protein, peptide, antibody, nanobody, or custom binder candidates for a target. Not for screening existing proteins or small molecules.

69

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Critical

Do not install without reviewing

SKILL.md
Quality
Evals
Security

Quality

Content

85%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 highly actionable, well-gated workflow document: executable commands, explicit spending and validation checkpoints, and clean one-level-deep disclosure into real reference files. Its single weakness is redundancy — several Always Do This bullets restate rules already given in the workflow steps or Run sizing section.

Suggestions

Consolidate the Codex background/heartbeat guidance: step 10 and Always Do This bullets on download-results/heartbeat/never-manual-poll overlap heavily — state the per-runtime pattern once and cross-reference it.

Fold the Always Do This bullet repeating the exploration-first mandate into a one-line pointer to step 1, since the full rule is already there.

State the num_proteins range and default tier once (Run sizing) and reference it from step 6 and the Always Do This bullet instead of restating both.

DimensionReasoningScore

Conciseness

The body is free of concept padding and dense with operational specifics, but key rules are repeated: the exploration-first mandate appears in step 1 and again in the first Always Do This bullet; the spending gate is explained in step 8, in Run sizing, and in Always Do This; Codex background/heartbeat handling spans step 10 plus three Always Do This bullets. This is more than the minor over-explanation of anchor 4 but well short of anchor 2's concept padding.

3 / 5

Actionability

The Command Pattern block gives copy-paste-ready boltz-api commands with concrete flag forms (@yaml:///absolute/path/payload.yaml, --raw-output --transform id), payload key names are stated, and the sequence DSL is illustrated with exact examples ("ACDE8GHI", "MKTAYI5..10VKSHFSRQ"), covering the common cases.

5 / 5

Workflow Clarity

An 11-step sequence with explicit validation checkpoints for a money-spending batch operation: estimate-cost plus a fresh-approval spending gate before every start, range enforcement of 10 <= num_proteins <= 1,000,000 before costing, install/auth retry loops at the top, and a defined ranking step. This matches the anchor with explicit validation, feedback loops, and a checklist section.

5 / 5

Progressive Disclosure

The body stays an overview and cleanly defers detail to real, one-level-deep, well-signaled references: api shapes/rules to references/api.md, output layout and metrics to references/results.md, and the exploration pass to references/target-exploration.md (which itself points to scripts/). Each reference carries a when-to-read cue, matching the clear-overview anchor.

5 / 5

Total

18

/

20

Passed

Description

82%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, third-person description with an explicit use-when trigger clause, natural modality keywords, and a negative boundary that sharply reduces conflict risk. Its only weakness is that it states one or two actions rather than a fuller picture of the workflow, which caps specificity.

DimensionReasoningScore

Specificity

"Design new protein binders" plus "generating protein, peptide, antibody, nanobody, or custom binder candidates" names the domain with 1-2 concrete actions, but does not list several actions (no mention of cost estimation, campaign sizing, or result ranking), matching the anchor for 1-2 concrete actions rather than the several-actions anchor at 4.

3 / 5

Completeness

It explicitly answers both what ("Design new protein binders with Boltz") and when ("Use when generating protein, peptide, antibody, nanobody, or custom binder candidates for a target"), with concrete trigger phrases and an additional explicit when-not clause ("Not for screening existing proteins or small molecules").

5 / 5

Trigger Term Quality

Natural user terms are present ("protein, peptide, antibody, nanobody, or custom binder", "design", "candidates for a target"), giving good coverage; common synonyms like VHH, Fab, miniprotein, or "de novo" are missing, so it sits at good-but-not-comprehensive coverage rather than 5 and clearly above the missing-variations anchor at 3.

4 / 5

Distinctiveness Conflict Risk

It carves a clear niche (de novo binder generation with Boltz) and the explicit exclusion "Not for screening existing proteins or small molecules" minimizes overlap with screening and small-molecule skills, matching the clear-niche/minimal-conflict anchor.

5 / 5

Total

17

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
openai/plugins
Reviewed

Table of Contents

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.