AB image-deck
image-deck creates PPT and PowerPoint presentations, slide decks, single slides, carousel pages, and full-image decks through Codex built-in image_gen (GPT Image 2). Trigger for requests to make PPT, PowerPoint, slides, a deck, presentation, or carousel, including 做PPT, 制作PPT, 帮我做PPT, 生成PPT, 做deck, 做slides, and 做演示文稿. Each page is one complete generated raster image with visible text inside it. Before planning, confirm page count, language, style, and text richness/content density; recommend page count, style, and text richness, but let the user choose language. Use two mandatory, distinct gates: overall design approval authorizes exactly one master sample; sample-style approval after the sample is displayed authorizes the remaining slides. Default to information-rich 图文并茂 pages unless the user chooses concise or image-led pages.
As a process B 70/100 · Nearly there — weak spots: inputs and preconditions, execution cost
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 6. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 11596 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 70/100
- 30Inputs and preconditions. Does not say what the process needs to start
- 40Execution cost. Instruction body is 11596 tokens: crowds the task out of the window
- 60Tools and files. Uses tools (python) that frontmatter does not declare
- 60Steps. 218 steps, 4 vague phrases
- 60Result and completion. Output format stated, no completion criterion
- 100When it triggers. States when to use and when not to
- 100Failures and branches. 28 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 17 top-level sections: this looks like several domains in one skill
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 841: 120–800 characters recommended
- +1No license
- +2Single-language instructions
- +4Structure: 30 headings
- +3Step-by-step instructions: 218 items
- +3Output format is stated explicitly
- +4Has examples (11 code blocks)
- +4Reference files are cited in the instructions (1 of 1)
- +3All 1 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 81.