CtrlK
BlogDocsLog inGet started
Tessl Logo

gpt-image2-ppt

Generate visually striking PPT slides via OpenAI's gpt-image-2 -- use any style in styles/<collection>/STYLE_ID.md or mimic a user-supplied .pptx template; outputs high-res slide PNGs and a 16:9 .pptx. Use when the user asks to make a presentation, slides, deck, pitch deck, investor PPT, magazine-style PPT, or 做一份 PPT / 生成幻灯片 / 用 gpt-image 生成 PPT / 按这个模板生成 PPT.

72

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Failed to scan

The risk profile of this skill

SKILL.md
Quality
Evals
Security

Quality

Content

81%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.

The body is highly actionable with excellent validation-gated workflows, but it is over-long for a SKILL.md: redundant render-backend sections and fully-detailed backend variants inflate token cost, and several referenced bundle directories (styles/, examples/, docs/) are not present in the bundle. Tightening redundancy and moving the large style/recipe tables into their referenced files would lift both conciseness and progressive disclosure.

Suggestions

Deduplicate the render-backend precheck: state it once and have the template-clone and editable-mode sections reference it, instead of restating the full platform/install table twice.

Move the 33-row styles table and 6-row recipes table into styles/README.md and examples/ respectively, leaving the SKILL.md body a concise pointer plus the top representative examples.

Ensure the referenced bundle directories (styles/, examples/, docs/) actually ship alongside scripts/, or trim body references to files that exist, so progressive-disclosure navigation resolves to real files.

DimensionReasoningScore

Conciseness

At 810 lines the body is mostly efficient operational guidance (no 'what is a PPT' padding), but it is noticeably long with redundancy — the render-backend precheck is restated verbatim in both the template-clone and editable-mode sections, and three backend paths are each explained in full; not 2 because the content is substantive rather than padded, not 4 because the redundancy could clearly be tightened.

3 / 5

Actionability

Fully executable, copy-paste-ready commands throughout — `generate_ppt.py --prepare-only`, `render_template.py --check`, `md_to_plan.py`, `--edit`/`--element-updates` examples, JSON slide_spec blocks, and a pptx packaging snippet — covering the API-direct, Codex-native, codex-backend, template-clone, editable, and external-image cases.

5 / 5

Workflow Clarity

Multi-step processes are explicitly numbered ('生成流程', 'Skill 调用规范', '生成流程(模板克隆)') with explicit validation checkpoints and feedback loops — '在跑任何 --template-pptx 命令之前...先检查', `render_template.py --check`, and per-page review of `editable_renders/page-XX.png` — so the destructive/batch cap does not apply.

5 / 5

Progressive Disclosure

References are well-signaled and one-level-deep ('完整结构...见 docs/workflow.md', 'docs/edit_guide.md', 'styles/README.md'), but scored against the actual bundle only `scripts/` exists while the referenced `styles/`, `examples/`, and `docs/` directories are missing, and the 33-row styles table and 6-row recipe table are inlined rather than split out; not 5 because of the missing bundle dirs and inlined reference-table content, not 3 because the in-body navigation is genuinely clear.

4 / 5

Total

17

/

20

Passed

Description

100%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.

The description is exemplary: third-person voice, concrete capabilities, comprehensive natural trigger terms (English + Chinese), and an explicit 'Use when...' clause. It cleanly answers what and when while occupying a distinct gpt-image-2 niche.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'Generate visually striking PPT slides', 'use any style in styles/<collection>/STYLE_ID.md', 'mimic a user-supplied .pptx template', 'outputs high-res slide PNGs and a 16:9 .pptx' — matching the comprehensive-coverage anchor; not 4 because coverage is broad rather than having minor gaps.

5 / 5

Completeness

Explicitly answers both 'what' (generate slides via gpt-image-2, use styles or mimic a template, output PNGs + .pptx) and 'when' ('Use when the user asks to make a presentation, slides, deck...') with concrete trigger phrases; not 4 because the 'when' is explicit and specific rather than weak.

5 / 5

Trigger Term Quality

Comprehensive natural terms including synonyms and language variants — 'presentation, slides, deck, pitch deck, investor PPT, magazine-style PPT' plus Chinese triggers '做一份 PPT / 生成幻灯片 / 用 gpt-image 生成 PPT / 按这个模板生成 PPT'; matches the comprehensive-synonyms anchor.

5 / 5

Distinctiveness Conflict Risk

Clear niche tied to 'OpenAI's gpt-image-2' image-based PPT generation with distinct, model-specific triggers; minimal overlap risk with general slide skills.

5 / 5

Total

20

/

20

Passed

Validation

87%

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

Validation14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (811 lines); consider splitting into references/ and linking

Warning

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

14

/

16

Passed

Repository
JuneYaooo/gpt-image2-ppt-skills
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.