image-gen

Generate images via Nano Banana (Gemini 2.5/3.1 Flash Image) on OpenRouter. Use when the user asks to draw, illustrate, render or generate any kind of picture/diagram/scene.

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

92%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.

Highly actionable, tight, and well-structured single-file skill with executable code and useful error-recovery guidance. The main improvement target is the time-sensitive cost/version figures, which could rot and would benefit from being framed as observed-at-writing-time.

Suggestions

Mark the "Cost (observed)" figures as a snapshot with the date, or move them to a deprecated/old-patterns framing, so stale numbers are not mistaken for current pricing.

Optionally split the full request/response shapes and troubleshooting into reference files if the skill grows, to preserve the lean quick-recipe overview.

Dimension	Reasoning	Score
Conciseness	Lean, code-first, and free of padding or explanation of basics Claude already knows; the only drag is the time-sensitive "Cost (observed)" section with dated token math and version numbers, which risks staleness without a deprecated/old-patterns framing.	4 / 5
Actionability	Fully executable copy-paste Python recipe, complete request/response JSON shapes, and a concrete image-input example covering the common generate and edit cases.	5 / 5
Workflow Clarity	Single-purpose skill with a clear quick recipe, and the Troubleshooting section supplies explicit error-recovery checkpoints (empty images, 403, refusal, corrupt PNG) that function as validation/feedback loops.	5 / 5
Progressive Disclosure	A self-contained single SKILL.md with no bundle files and well-organized sections (Quick recipe, Request/Response shape, Cost, When to use, Troubleshooting); per the under-50-line no-external-refs rule this is appropriately structured.	5 / 5
	Total	19 / 20 Passed

Description

95%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 nails the what/when structure with natural trigger terms and model specificity. The only minor gap is that the description omits edit/variation actions that the body actually supports.

Suggestions

Consider adding 'edit or vary an existing image' to the capability list in the description, since the skill body documents image-input editing.

Dimension	Reasoning	Score
Specificity	Names the domain and concrete action ("Generate images") plus specific models/provider ("Nano Banana (Gemini 2.5/3.1 Flash Image) on OpenRouter"), but does not enumerate edit/extend/variate actions in the description, leaving a minor coverage gap.	4 / 5
Completeness	Explicitly answers both what ("Generate images via Nano Banana ... on OpenRouter") and when ("Use when the user asks to draw, illustrate, render or generate...") with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive natural verb coverage — "draw, illustrate, render or generate" — plus concrete object nouns ("picture/diagram/scene") that users would naturally say.	5 / 5
Distinctiveness Conflict Risk	Clear niche (generative image creation via a specific model family on a specific provider) with distinct triggers and minimal overlap risk with plotting/diagram skills.	5 / 5
	Total	19 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata' field is not a dictionary	Warning
metadata_field	'metadata' field is not a dictionary	Warning

	Total	14 / 16 Passed

Repository: EverMind-AI/Raven
Path: demos/skill_retrieval/skills/image-gen/SKILL.md
Commit: cd68645

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.