paper-illustration-image2

Generate publication-quality academic illustrations through a local Codex app-server bridge that uses Codex native image generation. This is a separate experimental alternative to `paper-illustration`, intended for Claude Code users who want a GPT-image-style renderer without modifying the original skill.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable and the multi-stage workflow is clearly sequenced with real validation checkpoints, which are its strengths. Its weaknesses are verbosity (redundant sections and an oversized inline style guide/diagram) and a broken external reference under uneven progressive disclosure.

Suggestions

Move the CVPR/ICLR/NeurIPS style guide into a references/ file (e.g. STYLE.md) and replace the inline block with a one-level-deep, clearly signaled link.

Consolidate the overlapping "Key Rules", "Repair Path", and "Model Summary" sections to remove restated content.

Fix or remove the dangling reference to ../shared-references/integration-contract.md so all cited paths resolve to real bundle files.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient but padded in places: a large ASCII workflow diagram, a full style guide with non-English commentary, verbose resolver-layer bash, and overlapping "Key Rules", "Repair Path", and "Model Summary" sections that restate prior content.	2 / 3
Actionability	Concrete, copy-paste-ready commands with real flags are given throughout (preflight, generate_start/generate_status with timeoutSeconds, finalize with --best-image/--score, verify with --json-out), matching the anchor for fully executable guidance.	3 / 3
Workflow Clarity	A clear Step 0–7 sequence is present with explicit validation checkpoints ("Confirm preflight JSON says ok=true before rendering"), a scored feedback loop, and a reject-until-≥9 acceptance gate, matching the anchor for clear sequencing with validation.	3 / 3
Progressive Disclosure	Only one bundle file exists (scripts/paper_illustration_image2.py) with no references/ or assets/, yet the body keeps a full inline style guide and large ASCII diagram that could be split out, and it references ../shared-references/integration-contract.md which does not exist in the bundle.	2 / 3
	Total	10 / 12 Passed

Description

67%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description clearly states what the skill does and how it differs from `paper-illustration`, giving it strong specificity and distinctiveness. Its weakness is the missing explicit "Use when..." trigger clause, which caps both completeness and trigger-term quality at 2.

Suggestions

Append an explicit trigger clause, e.g. ".Use when generating architecture diagrams, method illustrations, or pipeline figures for a paper."

Add common natural trigger terms a user would actually say, such as "figure", "diagram", "architecture diagram", or "paper figure".

Dimension	Reasoning	Score
Specificity	"Generate publication-quality academic illustrations through a local Codex app-server bridge that uses Codex native image generation" names multiple concrete actions and the exact mechanism, matching the anchor for listing specific concrete actions.	3 / 3
Completeness	The "what" is clearly answered, but "when" is only implied via "intended for Claude Code users who want..." with no explicit "Use when..." clause, which per the guidelines caps completeness at 2.	2 / 3
Trigger Term Quality	Terms like "academic illustrations", "paper", and "GPT-image-style renderer" are relevant, but common natural variations a user would actually say (e.g. "figure", "diagram", "architecture diagram") are under-covered and there is no explicit trigger clause.	2 / 3
Distinctiveness Conflict Risk	"This is a separate experimental alternative to `paper-illustration`" carves out a clear niche and explicitly differentiates it from the sibling skill, making wrong-skill triggering unlikely.	3 / 3
	Total	10 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 suspicious	Warning

	Total	13 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Commit: 82076e5

Reviewed: 5 days 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.