chai1

Run or prepare Chai-1 structure predictions for biomolecular complexes. Use when a task asks for Chai-1 inputs, multimers, ligand/nucleic acid structure prediction, or confidence review.

Quality

79%

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

Fix and improve this skill with Tessl

tessl review fix ./skills/chai1/SKILL.md

Quality

Content

75%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 concise, well-sequenced workflow overview that assumes Claude's competence and includes useful verification checkpoints. Its main limitation is actionability: it describes what to do at each step but rarely provides the executable commands, examples, or manifest format needed to carry it out.

Suggestions

Add a concrete input-manifest example and a sample Chai-1 invocation (CLI command or API request body) for at least one execution path so the guidance is executable, not just descriptive.

Provide runnable commands or code snippets for the common execution paths (e.g. local model CLI, Modal job) instead of only listing the path names in step 2.

Add an explicit error-recovery loop, e.g. 'If interface confidence is low or stereochemistry warnings appear, re-run with adjusted restraints/templates and re-check before using the structure as evidence.'

Dimension	Reasoning	Score
Conciseness	Lean 5-step workflow with no padding and no explanation of concepts Claude already knows; every line is actionable guidance and the single framing sentence earns its place. Matches the lean-and-efficient top anchor.	5 / 5
Actionability	Provides a concrete checklist (manifest components, capture list, review criteria) but no executable commands, code, manifest-format example, or sample invocation for any execution path. Not 4 because key executable details are missing; not 2 because the guidance is more structured than a bare high-level hint.	3 / 5
Workflow Clarity	Clear 5-step sequence with a verify-before-running checkpoint (step 2), a review step (step 5), and a final gate ('Use the structure as evidence only after the provenance and confidence checks are attached'). Not 5 because there is no explicit error-recovery/retry loop for low-confidence or warning cases.	4 / 5
Progressive Disclosure	Short, well-organized body with a clear numbered workflow and no nested references, fitting the simple-skill pattern. Not 5 because the skill spans five execution paths and a complex input manifest that could reasonably warrant reference files, and none are signaled.	4 / 5
	Total	16 / 20 Passed

Description

83%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, concise description that clearly states what the skill does and when to use it with concrete, domain-appropriate trigger terms. The main weakness is minor overlap risk from generic structure-prediction triggers that are not always qualified with 'Chai-1'.

Dimension	Reasoning	Score
Specificity	Names the Chai-1 domain plus several concrete actions ('Run or prepare Chai-1 structure predictions') and trigger capabilities (multimers, ligand/nucleic acid prediction, confidence review), with only minor coverage gaps. Not 5 because 'run or prepare' is somewhat generic and template/restraint setup actions are implied rather than listed.	4 / 5
Completeness	Explicitly answers both what ('Run or prepare Chai-1 structure predictions for biomolecular complexes') and when ('Use when a task asks for Chai-1 inputs, multimers, ligand/nucleic acid structure prediction, or confidence review') with concrete trigger phrases. Clearly matches the top anchor.	5 / 5
Trigger Term Quality	Includes natural domain terms a user would say ('Chai-1 inputs', 'multimers', 'ligand/nucleic acid structure prediction', 'confidence review'). Not 5 because synonyms (e.g. protein complex, docking) and file extensions (.pdb, .cif) are absent.	4 / 5
Distinctiveness Conflict Risk	The 'Chai-1' branding gives a clear niche, but the unqualified triggers 'multimers, ligand/nucleic acid structure prediction, or confidence review' could overlap with related structure-prediction skills (AlphaFold, ColabFold). Not 5 because of this minor overlap risk.	4 / 5
	Total	17 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: companion-inc/feynman
Path: skills/chai1/SKILL.md
Commit: 95c51d9

Reviewed: about 10 hours 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.