CommunityImagegithub.com

Abaans/abaans-agent-skills

Research, generate and refine logo concepts through explicit lettering structure, visual references and focused critique. Use for wordmarks, monograms, emblems, logo prompts or concept reviews; generated images remain concept studies.

What is abaans-agent-skills?

abaans-agent-skills is a Claude Code agent skill that research, generate and refine logo concepts through explicit lettering structure, visual references and focused critique. Use for wordmarks, monograms, emblems, logo prompts or concept reviews; generated images remain concept studies.

Works with~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/Abaans/abaans-agent-skills/tree/HEAD/skills/logo-concept-generation

Installed? Explore more Image skills: steipete/songsee, affaan-m/frontend-design, mattpocock/improve-codebase-architecture · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

Logo concept generation

Turn a loose visual direction into distinct logo concepts that can be evaluated and redrawn as vectors. Match the requested stage: research, prompts, generation, critique or refinement. A prompts-only request does not authorize image generation.

Requirements

Use available web or visual-reference tools for research, image-generation tools for requested images, and image inspection for critique. Use the host's supported image workflow and its approval requirements. If a capability is unavailable, complete the stages that are possible and name the missing capability.

Establish the brief

Inspect supplied briefs, design files and references first. Resolve the exact brand spelling, intended audience, desired impression, uses, palette, existing elements to preserve and prohibited motifs. Ask only for missing choices that materially change the result.

Replace ambiguous language such as “typographic logo” with a specific structure: wordmark, connected custom lettering, monogram, lettermark, emblem or combination mark. Explain which letters connect, overlap or share strokes. Keep approved source artwork intact.

Research and choose directions

When research is requested, move from the broad tone to the brand category, an adjacent category, the desired lettering mechanism and small monochrome examples. Stop when references repeat the same idea. Record source links, useful attributes and features to avoid. Transfer principles such as rhythm, contrast or spacing; do not copy an existing mark.

Create a small number of structurally distinct directions appropriate to the brief. Each needs a thesis, logo anatomy, geometric mechanism, exclusions and its main legibility or reproduction risk. A font change alone is rarely a distinct concept.

Read prompt and critique patterns when writing prompts, evaluating an image or preparing a vector handoff.

Generate and inspect

Write one committed construction per prompt. Include exact text, glyph relationships, composition, palette and exclusions. Remove either/or choices before generation. Start in flat monochrome unless the requested application depends on color. Use the supported mechanism to attach reference images.

Inspect every delivered image for spelling, letter count, intended connections, open counters, legibility, balance, spacing and silhouette. Reject attractive results that violate the structural brief. A raster that resembles a vector is still a raster.

Refine with explicit Keep / Change / Avoid instructions. Change one or two related variables at a time; describe geometry rather than asking only for “more premium.” Reuse the current image as an edit reference where supported. Stop once the concept is clear enough for vector construction.

Deliver the requested stage

  • Research: reference synthesis, sources and direction opportunities.
  • Prompts: direction cards and generation-ready prompts.
  • Generation: concept images, short critique and the next useful refinement.
  • Review: material issues and specific changes, without generating replacements unless requested.

Label concepts accurately. Production identity work still requires editable vector construction, optical spacing, small-size/reversed-color checks and any appropriate brand clearance. Visual research does not establish legal uniqueness.

See examples and review cases for stage-specific requests.

Individual skills in this repo

This repo contains 3 individual skills — each has its own dedicated page.

Related Skills