Route Selection
| Route |
Trigger |
Route File |
| HTML (default) |
All PDF creation requests |
routes/html.md |
| LaTeX |
User explicitly requests LaTeX, .tex, or Tectonic |
routes/latex.md |
| Process |
Work with existing PDFs (extract, merge, fill forms, etc.) |
routes/process.md |
Default to HTML. Only use LaTeX route when user explicitly requests it.
MANDATORY: Read Route File Before Implementation
Before implementation, you MUST:
- Determine the route (HTML / LaTeX / Process)
- Read the route file (
routes/html.md, routes/latex.md, or routes/process.md)
- Only then proceed with implementation
This file (SKILL.md) contains constraints and principles. Route files contain how-to details.
Decision Rules
Route Selection
| User Says |
Route |
| "Create a PDF", "Make a report", "Write a paper" |
HTML |
| "Use LaTeX", "Compile .tex", "Use Tectonic" |
LaTeX |
| "Extract text from PDF", "Merge these PDFs", "Fill this form" |
Process |
Cover Style Selection (HTML Route)
| Context |
Style |
| Academic paper, thesis, formal coursework |
Minimal (white, centered, no decoration) |
| Reports, proposals, professional documents |
Designed (choose from style reference in html.md) |
| Uncertain |
Default to Designed — plain text cover = mediocre |
Key principle: Cover background separates "acceptable" from "impressive". See html.md for 11 style options.
Citation Format Selection
| Document Language |
Format |
| Chinese |
GB/T 7714 (use [J][M][D] identifiers) |
| English |
APA |
| Mixed |
Chinese refs → GB/T 7714, English refs → APA |
Quick Start
Use the unified CLI for all operations:
# Check environment (JSON output, exit code 0=ok, 2=missing deps)
/app/.kimi/skills/kimi-pdf/scripts/pdf.sh check
# Auto-fix missing dependencies (idempotent, safe to run multiple times)
/app/.kimi/skills/kimi-pdf/scripts/pdf.sh fix
# Convert HTML to PDF
/app/.kimi/skills/kimi-pdf/scripts/pdf.sh html input.html
# Compile LaTeX to PDF
/app/.kimi/skills/kimi-pdf/scripts/pdf.sh latex input.tex
Exit codes:
0 = success
1 = usage error
2 = dependency missing (run pdf.sh fix)
3 = runtime error
Dependencies by route:
- HTML route: Node.js, Playwright, Chromium
- Process route: Python 3, pikepdf, pdfplumber
- LaTeX route: Tectonic
Core Constraints (Must Follow)
1. Output Language
Output language must match user's query language.
- User writes in Chinese → PDF content in Chinese
- User writes in English → PDF content in English
- User explicitly specifies language → Follow user's specification
2. Word Count and Page Constraints
- Strictly follow user-specified word/page count requirements
- Do not arbitrarily inflate content length
3. Citation and Search Standards
CRITICAL: Search Before Writing
DO NOT fabricate information. When in doubt, SEARCH.
If content involves ANY of these, you MUST search FIRST before writing:
- Statistics, numbers, percentages, rankings
- Policies, regulations, laws, standards
- Academic research, theories, methodologies
- Current events, recent developments
- Anything you're not 100% certain about
When Search is Required
| Scenario |
Search? |
Notes |
| Statistics, data |
Required |
e.g., "2024 employment rate" |
| Policies, regulations |
Required |
e.g., "startup subsidies" |
| Research, papers |
Required |
e.g., "effectiveness of method X" |
| Time-sensitive content |
Required |
Information after knowledge cutoff |
| Uncertain facts |
Required |
If unsure, always search |
| Common knowledge |
Not needed |
e.g., "water boils at 100°C" |
Search workflow:
- Identify facts/data requiring verification
- Search for authentic sources
- If results insufficient, iterate search until reliable info obtained
- Include real sources in references
- If search fails repeatedly, tell the user instead of making up data
Citations Must Be Real
Fabricating references is prohibited. All citations must have:
- Correct author/institution names
- Accurate titles
- Verifiable year, journal/source
Cross-references (Must Be Clickable)
As shown in <a href="#fig-1-1">Figure 1-1</a>...
From <a href="#eq-2-1">Equation (2-1)</a>...
See <a href="#sec3">Section 3</a>...
Note: id must be placed at container top (see CSS Counters section in html.md).
Content Quality Constraints
1. Word/Page Count Constraints
Must strictly follow user-specified word or page count requirements:
| User Request |
Execution Standard |
| Specific word count (e.g., "3000 words") |
Within ±20%, i.e., 2400-3600 words |
| Specific page count (e.g., "5 pages") |
Exactly equal, last page may be partial |
| Word count range (e.g., "2000-3000 words") |
Must fall within range |
| No explicit requirement |
Infer reasonably by document type; prefer thorough over superficial |
| Minimum specified (e.g., "more than 5000 words") |
No more than 2x, i.e., 5000-10000 words |
Prohibited behaviors:
- Arbitrarily shortening content ("concise" is not an excuse)
- Padding pages with excessive bullet lists (maintain high information density)
- Exceeding twice the user's requested length
Special case - Resume/CV:
- Resume should be 1 page unless user specifies otherwise
- Use compact margins:
margin: 1.5cm
2. Outline Adherence (Mandatory)
When user provides outline:
- Strictly follow the user-provided outline structure
- Section titles must match outline (minor wording adjustments OK, no level/order changes)
- Do not add or remove sections arbitrarily
- If outline seems problematic, ask user first before modifying
When no user outline:
- Use standard structures based on document type:
- Academic papers: IMRaD (Introduction-Methods-Results-Discussion) or Introduction-Literature Review-Methods-Results-Discussion-Conclusion
- Business reports: Conclusion-first (Executive Summary → Detailed Analysis → Recommendations)
- Technical docs: Overview → Principles → Usage → Examples → FAQ
- Course assignments: Follow assignment structure requirements
- Sections must have logical progression, no disconnects
Tech Stack Overview
| Route |
Tools |
Purpose |
| HTML |
Playwright + Paged.js |
HTML → PDF conversion |
| HTML |
KaTeX, Mermaid |
Math formulas, diagrams |
| Process |
pikepdf |
Form filling, page operations, metadata |
| Process |
pdfplumber |
Text and table extraction |
| Process |
LibreOffice |
Office → PDF conversion |
| LaTeX |
Tectonic |
LaTeX → PDF compilation |
1---
2name: kimi-pdf
3description: Professional PDF solution. Create PDFs using HTML+Paged.js (academic papers, reports, documents). Process existing PDFs using Python (read, extract, merge, split, fill forms). Supports KaTeX math formulas, Mermaid diagrams, three-line tables, citations, and other academic elements. Also use this skill when user explicitly requests LaTeX (.tex) or native LaTeX compilation.
4---56
7## Route Selection
8
9| Route | Trigger | Route File |
10|-------|---------|------------|
11| **HTML** (default) | All PDF creation requests | `routes/html.md` |
12| **LaTeX** | User explicitly requests LaTeX, .tex, or Tectonic | `routes/latex.md` |
13| **Process** | Work with existing PDFs (extract, merge, fill forms, etc.) | `routes/process.md` |
14
15**Default to HTML.** Only use LaTeX route when user explicitly requests it.
16
17### MANDATORY: Read Route File Before Implementation
18
19<system-reminder>
20You MUST read the corresponding route file before writing ANY code.
21Route files contain critical implementation details NOT duplicated here.
22Skipping this step leads to incorrect output (wrong scripts, missing CSS, broken layouts).
23</system-reminder>
24
25**Before implementation, you MUST:**
261. Determine the route (HTML / LaTeX / Process)
272. **Read the route file** (`routes/html.md`, `routes/latex.md`, or `routes/process.md`)
283. Only then proceed with implementation
29
30This file (SKILL.md) contains constraints and principles. Route files contain **how-to details**.
31
32### Decision Rules
33
34#### Route Selection
35| User Says | Route |
36|-----------|-------|
37| "Create a PDF", "Make a report", "Write a paper" | HTML |
38| "Use LaTeX", "Compile .tex", "Use Tectonic" | LaTeX |
39| "Extract text from PDF", "Merge these PDFs", "Fill this form" | Process |
40
41#### Cover Style Selection (HTML Route)
42| Context | Style |
43|---------|-------|
44| Academic paper, thesis, formal coursework | **Minimal** (white, centered, no decoration) |
45| Reports, proposals, professional documents | **Designed** (choose from style reference in html.md) |
46| Uncertain | Default to **Designed** — plain text cover = mediocre |
47
48**Key principle**: Cover background separates "acceptable" from "impressive". See html.md for 11 style options.
49
50#### Citation Format Selection
51| Document Language | Format |
52|-------------------|--------|
53| Chinese | GB/T 7714 (use [J][M][D] identifiers) |
54| English | APA |
55| Mixed | Chinese refs → GB/T 7714, English refs → APA |
56
57---
58
59## Quick Start
60
61**Use the unified CLI for all operations:**
62
63```bash
64# Check environment (JSON output, exit code 0=ok, 2=missing deps)
65/app/.kimi/skills/kimi-pdf/scripts/pdf.sh check
66
67# Auto-fix missing dependencies (idempotent, safe to run multiple times)
68/app/.kimi/skills/kimi-pdf/scripts/pdf.sh fix
69
70# Convert HTML to PDF
71/app/.kimi/skills/kimi-pdf/scripts/pdf.sh html input.html
72
73# Compile LaTeX to PDF
74/app/.kimi/skills/kimi-pdf/scripts/pdf.sh latex input.tex
75```
76
77**Exit codes:**
78- `0` = success
79- `1` = usage error
80- `2` = dependency missing (run `pdf.sh fix`)
81- `3` = runtime error
82
83**Dependencies by route:**
84- **HTML route**: Node.js, Playwright, Chromium
85- **Process route**: Python 3, pikepdf, pdfplumber
86- **LaTeX route**: Tectonic
87
88---
89
90## Core Constraints (Must Follow)
91
92### 1. Output Language
93**Output language must match user's query language.**
94- User writes in Chinese → PDF content in Chinese
95- User writes in English → PDF content in English
96- User explicitly specifies language → Follow user's specification
97
98### 2. Word Count and Page Constraints
99- Strictly follow user-specified word/page count requirements
100- Do not arbitrarily inflate content length
101
102### 3. Citation and Search Standards
103
104#### CRITICAL: Search Before Writing
105**DO NOT fabricate information. When in doubt, SEARCH.**
106
107If content involves ANY of these, you **MUST search FIRST** before writing:
108- Statistics, numbers, percentages, rankings
109- Policies, regulations, laws, standards
110- Academic research, theories, methodologies
111- Current events, recent developments
112- **Anything you're not 100% certain about**
113
114<system-reminder>
115Never proceed with writing if you need statistics, research data, or policy information without searching first.
116Making up facts is strictly prohibited. When uncertain, search.
117</system-reminder>
118
119#### When Search is Required
120| Scenario | Search? | Notes |
121|----------|---------|-------|
122| Statistics, data | **Required** | e.g., "2024 employment rate" |
123| Policies, regulations | **Required** | e.g., "startup subsidies" |
124| Research, papers | **Required** | e.g., "effectiveness of method X" |
125| Time-sensitive content | **Required** | Information after knowledge cutoff |
126| **Uncertain facts** | **Required** | If unsure, always search |
127| Common knowledge | Not needed | e.g., "water boils at 100°C" |
128
129**Search workflow**:
1301. Identify facts/data requiring verification
1312. Search for authentic sources
1323. If results insufficient, **iterate search** until reliable info obtained
1334. Include real sources in references
1345. **If search fails repeatedly, tell the user** instead of making up data
135
136#### Citations Must Be Real
137**Fabricating references is prohibited**. All citations must have:
138- Correct author/institution names
139- Accurate titles
140- Verifiable year, journal/source
141
142#### Cross-references (Must Be Clickable)
143```html
144As shown in <a href="#fig-1-1">Figure 1-1</a>...
145From <a href="#eq-2-1">Equation (2-1)</a>...
146See <a href="#sec3">Section 3</a>...
147```
148**Note**: `id` must be placed at container top (see CSS Counters section in html.md).
149
150---
151
152## Content Quality Constraints
153
154### 1. Word/Page Count Constraints
155**Must strictly follow user-specified word or page count requirements**:
156
157| User Request | Execution Standard |
158|--------------|-------------------|
159| Specific word count (e.g., "3000 words") | Within ±20%, i.e., 2400-3600 words |
160| Specific page count (e.g., "5 pages") | Exactly equal, last page may be partial |
161| Word count range (e.g., "2000-3000 words") | Must fall within range |
162| No explicit requirement | Infer reasonably by document type; prefer thorough over superficial |
163| Minimum specified (e.g., "more than 5000 words") | No more than 2x, i.e., 5000-10000 words |
164
165**Prohibited behaviors**:
166- Arbitrarily shortening content ("concise" is not an excuse)
167- Padding pages with excessive bullet lists (maintain high information density)
168- Exceeding twice the user's requested length
169
170**Special case - Resume/CV**:
171- Resume should be **1 page** unless user specifies otherwise
172- Use compact margins: `margin: 1.5cm`
173
174### 2. Outline Adherence (Mandatory)
175**When user provides outline**:
176- **Strictly follow** the user-provided outline structure
177- Section titles must match outline (minor wording adjustments OK, no level/order changes)
178- Do not add or remove sections arbitrarily
179- If outline seems problematic, **ask user first** before modifying
180
181**When no user outline**:
182- Use standard structures based on document type:
183 - **Academic papers**: IMRaD (Introduction-Methods-Results-Discussion) or Introduction-Literature Review-Methods-Results-Discussion-Conclusion
184 - **Business reports**: Conclusion-first (Executive Summary → Detailed Analysis → Recommendations)
185 - **Technical docs**: Overview → Principles → Usage → Examples → FAQ
186 - **Course assignments**: Follow assignment structure requirements
187- Sections must have logical progression, no disconnects
188
189---
190
191## Tech Stack Overview
192
193| Route | Tools | Purpose |
194|-------|-------|---------|
195| HTML | Playwright + Paged.js | HTML → PDF conversion |
196| HTML | KaTeX, Mermaid | Math formulas, diagrams |
197| Process | pikepdf | Form filling, page operations, metadata |
198| Process | pdfplumber | Text and table extraction |
199| Process | LibreOffice | Office → PDF conversion |
200| LaTeX | Tectonic | LaTeX → PDF compilation |