Tessl • The management layer for AI agent skills

SkillPublic

shinka-setup

Create ShinkaEvolve task scaffolds from a target directory and task description, producing `evaluate.py` and `initial.<ext>` (multi-language). Use when asked to set up new ShinkaEvolve tasks, evaluation harnesses, or baseline programs for ShinkaEvolve.

Quality

80%

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/shinka-setup/SKILL.md

Quality

Content

67%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 thorough, actionable skill body with strong templates and a validation-backed workflow. Its main weaknesses are non-essential background prose that inflates token cost and a couple of navigation/path gaps in the optional bundle references.

Suggestions

Remove or compress the 'What is ShinkaEvolve?' and 'Evolution Flow' background sections to lean operational guidance; keep only the repo/paper links.

Fix the optional-template reference paths from 'skills/shinka-setup/scripts/run_evo.py' to the skill-relative 'scripts/run_evo.py' so navigation matches the actual bundle layout.

Either complete the non-Python evaluate.py template with a real subprocess call or explicitly justify why it remains a skeleton.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient with useful templates, but whole background sections ('What is ShinkaEvolve?', 'Evolution Flow', and the 'optimize for performance and discover novel scientific insights' goal prose) explain framework mechanics that are not needed to scaffold a task, going beyond the 'minor instances' threshold of anchor 4.	3 / 5
Actionability	Provides copy-paste-ready Python templates for initial.<ext> and evaluate.py plus a concrete smoke-test command, but the non-Python evaluate.py template is a placeholder skeleton ('# 1) Execute candidate program_path') and the optional run_evo.py/shinka.yaml are referenced rather than inlined, leaving minor gaps short of anchor 5.	4 / 5
Workflow Clarity	A clearly sequenced 7-step workflow with an explicit validation checkpoint (step 6: smoke-test command + required metrics/correct.json field checklist); it stops short of anchor 5 because there is no explicit fix-and-retry error-recovery loop. Validation is present, so the destructive/batch cap does not apply.	4 / 5
Progressive Disclosure	Clear section headers and signaled 'See ...' references to real bundle files (scripts/run_evo.py, scripts/shinka.yaml), but the reference paths use a 'skills/shinka-setup/scripts/...' prefix that does not match the actual './scripts/' bundle layout, and the two large code templates are inlined rather than split out.	4 / 5
	Total	15 / 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 well-constructed description with a clear what, an explicit 'Use when' trigger clause, and concrete named artifacts. It is held back from a perfect score by generic sub-trigger terms and a lack of synonym/extension coverage.

Suggestions

Add file-extension and synonym triggers (e.g., 'ShinkaEvolve, Shinka, evolutionary code optimization, evaluate.py, initial.py') to raise trigger-term coverage.

Tighten the 'when' clause so generic terms like 'evaluation harnesses' are always qualified by ShinkaEvolve to reduce overlap with general ML benchmarking skills.

Dimension	Reasoning	Score
Specificity	Names concrete artifacts (`evaluate.py`, `initial.<ext>`) and inputs (target directory, task description) with multi-language note, but the action is essentially one scaffold-creation task decomposed into outputs rather than the multiple distinct actions of the anchor-5 example.	4 / 5
Completeness	Clearly states what it does ('Create ShinkaEvolve task scaffolds...producing evaluate.py and initial.<ext>') and gives an explicit 'Use when...' clause with concrete trigger phrases, matching the anchor-5 example.	5 / 5
Trigger Term Quality	Includes natural phrases users would say ('set up new ShinkaEvolve tasks', 'evaluation harnesses', 'baseline programs'), but lacks synonyms and file-extension triggers needed for anchor 5.	4 / 5
Distinctiveness Conflict Risk	The repeated ShinkaEvolve framework name anchors a clear niche, but the sub-triggers 'evaluation harnesses' and 'baseline programs' are generic enough to create minor overlap risk with other ML eval/benchmark skills.	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: SakanaAI/ShinkaEvolve
Path: skills/shinka-setup/SKILL.md
Revision: 8adc053

Reviewed: 2 months ago

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.

shinka-setup

Table of Contents

Is this your skill?