skill-deck
- Repo stars 435
- Author updated Live
- Author repo nvim
- Domain
- Other
- Compatible agents
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- Trust score
- 88 / 100 · community maintained
- Author / version / license
- @benbrastmckie · no license declared
- Token usage
- Lean
- Setup complexity
- Guided setup
- External API key
- Not required
- Operating systems
- Unspecified (assume cross-platform)
- Runtime requirements
- No special requirements
- Permissions
-
- Read-only
- Write / modify
- Shell exec
- Network behavior
- Local-only
- Install commands
- 26 variants
Profile is derived at build time from SKILL.md and install vectors. Subject to drift from author intent.
Heads up: 未限定 allowed-tools,默认拥有全部工具权限。
---
name: skill-deck
description: Generate YC-style investor pitch decks in Typst Thin wrapper that routes pitch deck generation r…
category: other
runtime: no special runtime
---
# skill-deck output preview
## PART A: Task fit
- Use case: Generate YC-style investor pitch decks in Typst Thin wrapper that routes pitch deck generation requests to the deck-agent. Reference (do not load eagerly): This skill activates when: When an implementing agent encounters any of these patterns: Do not invoke for: runs entirely locally. Works with Claude Code, Cursor, Cline and 23 more..
- Inputs: target material, constraints, expected output, and acceptance criteria.
- Evidence boundary: follow “Context Pointers / Trigger Conditions / Direct Invocation” and do not present inference as author intent.
## PART B: Execution result
- **01** The card summarizes the use case; runtime output centers on “Generate YC-style investor pitch decks in Typst Thin wrapper that routes pitch deck generation requests to the deck-agent. Reference (do not load eagerly): This skill activates when: When an implementing agent encounters any of these patterns: Do not invoke for: runs entirely locally. Works with Claude Code, Cursor, Cline and 23 more.”.
- **02** When the source has headings, the agent prioritizes “Context Pointers / Trigger Conditions / Direct Invocation” so the result follows the author’s structure.
- **03** Typical output includes task judgment, concrete steps, required commands or file edits, validation, and follow-up options.
- **04** Risk context follows the fingerprint: read files, write/modify files, run shell commands; mostly runs locally; usually needs no extra API key.
## Running Rules
- read files, write/modify files, run shell commands; mostly runs locally; usually needs no extra API key.
- Validate with a small sample before expanding scope.
- Return the result, validation criteria, and next iteration options. The source mentions slash commands such as `/deck`; use them first when your agent supports command triggers.
Name target files or source material, expected output, forbidden changes, and whether network or shell access is allowed. Permission fingerprint: read files, write/modify files, run shell commands.
Start with a small task and check whether the result follows “Context Pointers / Trigger Conditions / Direct Invocation”. Inspect diffs, logs, previews, or tests before expanding scope.
Confirm the final output includes a concrete result, evidence, and next action. If it stays generic, tighten inputs, boundaries, and acceptance criteria.
---
name: skill-deck
description: Generate YC-style investor pitch decks in Typst Thin wrapper that routes pitch deck generation r…
category: other
source: benbrastmckie/nvim
---
# skill-deck
## When to use
- Generate YC-style investor pitch decks in Typst Thin wrapper that routes pitch deck generation requests to the deck-ag…
- Use it when the task has clear inputs, repeatable steps, and validation criteria.
## What to provide
- Target material, scope, expected result, and forbidden changes.
- Whether network, commands, file writes, or external services are allowed.
## Execution rules
- Organize steps around “Context Pointers / Trigger Conditions / Direct Invocation” and keep inference separate from source facts.
- read files, write/modify files, run shell commands; mostly runs locally; usually needs no extra API key.
- Validate with a small sample before expanding the task.
## Output requirements
- Return the deliverable, key evidence, validation method, and next action.
- Mark missing information as unknown; do not invent commands, platforms, or dependencies. The author source anchors workflow facts; repository files anchor sources and commands; Fluxly only adds fit, limitations, and quality judgment.
skill "skill-deck" {
input -> user goal + target files + boundaries + acceptance criteria
context -> Context Pointers / Trigger Conditions / Direct Invocation
rules -> SKILL.md triggers / order / output contract
runtime -> no special runtime | read files, write/modify files, run shell commands | mostly runs locally
guardrails -> usually needs no extra API key + small-sample validation + diff/log review
output -> copyable result + checklist + next iteration
} Deck Skill
Thin wrapper that routes pitch deck generation requests to the deck-agent.
Context Pointers
Reference (do not load eagerly):
- Path:
.opencode/context/core/formats/subagent-return.md - Purpose: Return validation
- Load at: Subagent execution only
Note: This skill is a thin wrapper. Context is loaded by the delegated agent, not this skill.
Trigger Conditions
This skill activates when:
Direct Invocation
- User explicitly runs
/deckcommand - User requests pitch deck generation in conversation
Implicit Invocation (during task implementation)
When an implementing agent encounters any of these patterns:
Plan step language patterns:
- "Generate pitch deck"
- "Create investor deck"
- "Build presentation for investors"
- "Make YC-style slides"
- "Create startup pitch"
Target mentions:
- "pitch deck"
- "investor presentation"
- "seed round slides"
- "fundraising deck"
- "touying pitch"
Task description keywords:
- "pitch deck generation"
- "investor slides"
- "startup presentation"
- "YC deck"
When NOT to trigger
Do not invoke for:
- Converting existing presentations (use skill-presentation)
- General slide conversion from PPTX (use skill-presentation)
- Document format conversions (use skill-filetypes)
- Creating non-investor presentation slides
Execution
1. Input Validation
Validate required inputs:
- At least one of:
prompt(text description) ORsource_path(file path) output_path- Optional, defaults based on inputtheme- Optional, defaults to "simple"slide_count- Optional, defaults to 10
# Validate at least one input source
if [ -z "$prompt" ] && [ -z "$source_path" ]; then
return error "Either a prompt or source file path is required"
fi
# If source_path provided, validate it exists
if [ -n "$source_path" ] && [ ! -f "$source_path" ]; then
return error "Source file not found: $source_path"
fi
# Convert source_path to absolute if relative
if [ -n "$source_path" ] && [[ "$source_path" != /* ]]; then
source_path="$(pwd)/$source_path"
fi
# Determine output path if not provided
if [ -z "$output_path" ]; then
if [ -n "$source_path" ]; then
source_dir=$(dirname "$source_path")
source_base=$(basename "$source_path" | sed 's/\.[^.]*$//')
output_path="${source_dir}/${source_base}-deck.typ"
else
output_path="$(pwd)/pitch-deck.typ"
fi
fi
# Convert output_path to absolute if relative
if [[ "$output_path" != /* ]]; then
output_path="$(pwd)/$output_path"
fi
2. Context Preparation
Prepare delegation context:
{
"prompt": "Startup description from user",
"source_path": "/absolute/path/to/startup-info.md",
"output_path": "/absolute/path/to/pitch-deck.typ",
"theme": "simple",
"slide_count": 10,
"metadata": {
"session_id": "sess_{timestamp}_{random}",
"delegation_depth": 1,
"delegation_path": ["orchestrator", "deck", "skill-deck"]
}
}
3. Invoke Agent
CRITICAL: You MUST use the Task tool to spawn the agent.
Required Tool Invocation:
Tool: Task (NOT Skill)
Parameters:
- subagent_type: "deck-agent"
- prompt: [Include prompt, source_path, output_path, theme, slide_count, metadata]
- description: "Generate pitch deck from input"
The agent will:
- Parse and validate inputs
- Read source file if provided
- Map content to YC's 10-slide structure
- Generate Typst code using touying
- Include speaker notes and TODO placeholders
- Return standardized JSON result
4. Return Validation
Validate return matches subagent-return.md schema:
- Status is one of: generated, partial, failed
- Summary is non-empty and <100 tokens
- Artifacts array present with output file path
- Metadata contains slide_count and slides_with_todos
5. Return Propagation
Return validated result to caller without modification.
Return Format
Expected successful return:
{
"status": "generated",
"summary": "Generated 10-slide pitch deck for Acme AI in Typst format. 2 slides have TODO placeholders.",
"artifacts": [
{
"type": "implementation",
"path": "/absolute/path/to/pitch-deck.typ",
"summary": "Touying pitch deck with speaker notes"
}
],
"metadata": {
"session_id": "sess_...",
"agent_type": "deck-agent",
"delegation_depth": 2,
"theme": "simple",
"slide_count": 10,
"slides_with_todos": 2,
"input_type": "prompt_and_file"
},
"next_steps": "Review and compile: typst compile pitch-deck.typ"
}
Error Handling
Input Validation Errors
Return immediately with failed status if neither prompt nor source file provided.
Source File Not Found
Return failed status with clear message about the missing file.
Agent Errors
Pass through the agent's error return verbatim.
Write Failure
Return failed status with directory/permission guidance.
Decide Fit First
Design Intent
How To Use It
Boundaries And Review