BF ppt-craft-editable
端到端 PPT 全流程，单技能自包含。用户首次触发本 skill 时，agent 必须默默跑 scripts/preflight.py 做环境自检 + 自动安装缺失的 Python 包；FAIL 才打断用户。 环境通过后走 conversation-first + image-first 工作流：多阶段对话、内容补强、风格预览、规划锁定、出每页定稿图，并在用户驱动下做图像级 retouch（去水印 / 去瑕疵，内置 IOPaint 自动安装）。最终交付高完成度图片型 PPTX。 Phase A 结束后必须主动询问用户是否要可编辑文字版（Phase C）；若同意，进入 Phase C：以 Phase A 定稿图作为视觉参考，重新生成"带装饰、留白文字区"的无字背景图 + HTML 编辑器调文字 → 一键渲染原生可编辑 PPTX。 用户明确要求"只做可编辑版 / 只跑 Phase C / 跳过图片版"时，直接进入 Phase C-only 模式，不执行 Phase A 的任何生成、预览、review、retouch。 ★ Phase D - PDF PPT编辑分支：用户上传 PDF 幻灯片时，自动判断每页是矢量文本还是图片型，用多模态模型抽取文案结构，在 HTML 预览页让用户确认后，生成无字背景 + 可编辑文字框，最终输出原生可编辑 PPTX。 ★ 当对话中出现 ===PPT-CRAFT-EDITABLE / DECK FROM EDITOR===、===PHASE-D / CONTENT EXTRACTION CONFIRMED===、旧标记 ===PPT-IMAGE-FIRST-EDITABLE / DECK FROM EDITOR=== 或 ===DECK JSON BEGIN/END=== 标记时，agent 必须立刻把标记之间的 JSON 原样写入对应文件并跑相应脚本，不解读不修改不追问。 当用户需要做汇报 / 答辩 / 路演 / 提案 PPT，或只丢一个主题想要完整成品，或上传 PDF 想转可编辑版时使用。
端到端 PPT 全流程，单技能自包含。用户首次触发本 skill 时，agent 必须默默跑 scripts/preflight.py 做环境自检 + 自动安装缺失的 Python 包；FAIL 才打断用户。 环境通过后走 conversation-first + image-first…
As a process F 36/100 · Will not run — References files that are not bundled: assets/.../index.html
How to improve
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- The text references files that are not there: add them or drop the references.
- 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 · 6
✓ No critical or high findings
Medium and low: 6
-
low Obfuscation
obf-base64-blobassets/candidate_picker_shell/build_candidate_picker_html.py:6Long base64-looking blob (quoted — discussed, not commanded)HTML_B64 = """PCFE…bWw+Cjxo…YWQ+CiAg…wIj
quoted -
low Secrets in code
secret-high-entropy-tokenassets/candidate_picker_shell/build_candidate_picker_html.py:6High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)HTML_B64 = """PCFE…bWw+Cjxo…YWQ+CiAg…wIj
quoted -
low Obfuscation
obf-base64-blobassets/preview_shell/build_preview_html.py:6Long base64-looking blob (quoted — discussed, not commanded)HTML_B64 = """PCFE…bWw+Cjxo…YWQ+CiAg…wIj
quoted -
low Secrets in code
secret-high-entropy-tokenassets/preview_shell/build_preview_html.py:6High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)HTML_B64 = """PCFE…bWw+Cjxo…YWQ+CiAg…wIj
quoted -
low Obfuscation
obf-base64-blobassets/review_shell/build_review_html.py:6Long base64-looking blob (quoted — discussed, not commanded)HTML_B64 = """PCFE…bWw+Cjxo…YWQ+CiAg…wIj
quoted -
low Secrets in code
secret-high-entropy-tokenassets/review_shell/build_review_html.py:6High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)HTML_B64 = """PCFE…bWw+Cjxo…YWQ+CiAg…wIj
quoted
Files scanned: 57. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
description-no-whendescription does not say WHEN to use the skill (no "use when") - warning
body-longSKILL.md body ≈ 6707 tokens (recommended < 5000); move details to references/ - warning
missing-refreference to a missing file: assets/.../index.html
Process rating: all ten parameters 36/100
- 0Tools and files. 1 referenced file(s) missing: assets/.../index.html
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 2 mutating operations with no state check
- 50Failures and branches. 0 branches, has a failure section
- 70Execution cost. Instruction body is 6707 tokens
- 100Steps. 182 steps
- 100Consistency. Name and required fields are in place
- 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
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 873: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- -242 emoji in the instructions: noise for the model
- -36 of 17 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +5Description quotes 2 example trigger phrases
- +4Structure: 53 headings
- +3Step-by-step instructions: 182 items
- +4Has examples (12 code blocks)
- +4Reference files are cited in the instructions (1 of 1)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 57.