CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

flux-best-practices

Comprehensive guide for BFL FLUX image generation models. Covers prompting, T2I, I2I, structured JSON, hex colors, typography, multi-reference editing, and model-specific best practices for FLUX.2 and FLUX.1 families.

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.claude/skills/flux-best-practices/SKILL.md

The canonical home for this skill is flux-best-practices in calesthio/OpenMontage

SKILL.md
Quality
Evals
Security

Quality

Content

65%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 well-organized, reasonably lean quick-reference with genuinely actionable prompting guidance, undermined by two structural issues: every external file it points to (AGENTS.md and all 11 rules/*.md) is missing from the bundle, and the prompt-composition workflow is never sequenced with a final verification step. Shipping the rule files (or inlining their essentials) and adding an explicit compose-then-verify sequence would lift both dimensions.

Suggestions

Ship the referenced bundle files or remove the links: all 11 rules/*.md files and AGENTS.md cited in the body are absent, so progressive disclosure currently dead-ends on every path.

Present the composition workflow as an explicit numbered sequence with a verification checkpoint, e.g., "1. Pick the model from the table. 2. Build the prompt from the formula. 3. Verify against the Critical Rules (no negative prompt, text quoted, lighting specified, hex colors paired with a description) before submitting."

Trim the AGENTS.md blockquote to a single clause (e.g., "See AGENTS.md for the extended upstream FLUX guide") — the repo-scoping and authority commentary is padding.

DimensionReasoningScore

Conciseness

The body is lean — the prompt formula, model table, six terse critical rules, and one worked example all earn their tokens and none of it re-explains what Claude already knows. The one trimmable spot is the AGENTS.md blockquote's repo-politics meta-commentary ("It is supplementary reference material scoped to this skill only... does not override or extend the repository-root AGENTS.md / AGENT_GUIDE.md").

4 / 5

Actionability

Concrete guidance throughout: the formula "[Subject] + [Action/Pose] + [Style/Medium] + [Context/Setting] + [Lighting] + [Camera/Technical]", a model table naming specific models ("FLUX.2 [flex] — Best text rendering"), "Use #RRGGBB format with color description", and a complete example prompt. Minor gaps: rules like hex colors and quoted typography lack inline mini-examples, deferring to rule files.

4 / 5

Workflow Clarity

The composition flow (model selection → formula → critical rules → example) is logically ordered but never presented as an explicit sequence, and there is no verification checkpoint such as "before finalizing, check the prompt against the Critical Rules". The skill is single-purpose but multi-part, so the simple-skill exception does not apply.

3 / 5

Progressive Disclosure

The in-body structure is good — 11 one-level-deep, clearly labeled rule links plus a flagged AGENTS.md reference — but the actual bundle contains no rules/ directory and no AGENTS.md, so every referenced path is broken and the disclosure architecture is nonfunctional as shipped. Good organization in the markdown itself keeps this above 2.

3 / 5

Total

14

/

20

Passed

Description

70%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 specific, well-scoped description with strong keyword coverage and minimal conflict risk, but it completely lacks a "Use when..." trigger clause, capping completeness at 3. Adding an explicit trigger (e.g., "Use when generating prompts or images with FLUX.2/FLUX.1 models") and spelling out T2I/I2I as text-to-image/image-to-image would raise it further.

Suggestions

Add a 'Use when...' trigger clause, e.g., "Use when generating images or writing prompts for FLUX.2 or FLUX.1 models, or when the user mentions FLUX, BFL, or text-to-image generation."

Spell out abbreviations as natural synonyms: "text-to-image (T2I)" and "image-to-image (I2I) editing" so users who don't use the jargon still match.

Convert topical nouns into concrete actions, e.g., "Write and optimize prompts... specify hex colors and quoted typography" instead of "Covers prompting, hex colors, typography".

DimensionReasoningScore

Specificity

"Covers prompting, T2I, I2I, structured JSON, hex colors, typography, multi-reference editing" enumerates many specific capabilities, but they are topical nouns rather than concrete actions and the skill's actual output (prompt-writing guidance) is only implied — minor gaps in coverage.

4 / 5

Completeness

The "what" is clear ("Comprehensive guide for BFL FLUX image generation models. Covers prompting, T2I, I2I..."), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the judging guidelines.

3 / 5

Trigger Term Quality

Natural terms like "FLUX", "image generation", "prompting", "typography", and "hex colors" are present, but spelled-out synonyms such as "text-to-image", "image-to-image", or "generate images" are missing and the abbreviations T2I/I2I are jargon-adjacent.

4 / 5

Distinctiveness Conflict Risk

"BFL FLUX image generation models" and "FLUX.2 and FLUX.1 families" define a precise vendor-model niche with distinct triggers; the only adjacency is a sibling bfl-api skill, which this prompting-focused description does not overlap.

5 / 5

Total

16

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 12 missing

Warning

Total

15

/

16

Passed

Repository

calesthio/OpenMontage

Path

.claude/skills/flux-best-practices/SKILL.md

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.