### /roast command example output schema

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Example JSON schema for the output of the /roast command, detailing the type of analysis, a title, the roast content, a pull quote, and the repository name.

```json
{
  "type": "roast",
  "title": "Gordon Ramsay on my-project",
  "content": "This codebase is so raw, it's still mooing. Your error handling strategy appears to be console.log and prayer. I've seen better architecture in a tent city.",
  "pullQuote": "Your error handling strategy appears to be console.log and prayer.",
  "repoName": "my-project"
}
```

--------------------------------

### Install Cursed Plugins via Claude Code

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Add and install the Cursed Plugins from the marketplace using the Claude Code CLI. Ensure you have Claude Code installed and configured.

```bash
# Add the plugin from marketplace
/plugin marketplace add Factory-AI/cursed-plugins

# Install the plugin
/plugin install cursed-plugins@cursed-plugins
```

--------------------------------

### Install Cursed Plugins via Factory Droid

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Add and install the Cursed Plugins from the marketplace using the Factory Droid CLI. Ensure you have Factory Droid installed and configured.

```bash
# Add the plugin from marketplace
droid plugin marketplace add https://github.com/Factory-AI/cursed-plugins

# Install the plugin
droid plugin install cursed-plugins@cursed-plugins
```

--------------------------------

### Install Cursed Plugins via Factory Droid

Source: https://github.com/factory-ai/cursed-plugins/blob/main/README.md

Use the droid CLI to add and install the plugin from the marketplace.

```bash
droid plugin marketplace add https://github.com/Factory-AI/cursed-plugins
droid plugin install cursed-plugins@cursed-plugins
```

--------------------------------

### /dating-profile command example output schema

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Example JSON schema for the output of the /dating-profile command, including the analysis type, a title, the dating profile content, a pull quote, and the repository name.

```json
{
  "type": "dating-profile",
  "title": "my-project's Dating Profile",
  "content": "I speak fluent TypeScript but mumble when anyone asks about my tests. Looking for someone who appreciates a monorepo lifestyle and doesn't mind 847 dependencies. My love language is barrel exports.",
  "pullQuote": "I speak fluent TypeScript but mumble when anyone asks about my tests.",
  "repoName": "my-project"
}
```

--------------------------------

### /agent-repellent command example output schema

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Example JSON schema for the output of the /agent-repellent command, showing the assessment type, title, content of the assessment, a pull quote, and the repository name.

```json
{
  "type": "agent-repellent",
  "title": "Agent Repellent Assessment: my-project",
  "content": "This codebase demonstrates exceptional resistance to AI comprehension. The variable naming convention appears to be a dialect known only to its author. Documentation exists in theory.",
  "pullQuote": "The variable naming convention appears to be a dialect known only to its author.",
  "repoName": "my-project"
}
```

--------------------------------

### Feng Shui: Analyze Structural Patterns

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

These are examples of structural patterns analyzed by the feng shui module, including directory layout, naming conventions, and dependency graphs.

```bash
# - Directory layout and nesting depth
# - File naming conventions
# - Import graph depth and circular dependencies
# - Utils/helpers/shared module placement
# - Barrel file patterns (index.ts re-exports)
# - Config file locations
# - Test and source file relationships
```

--------------------------------

### Install Cursed Plugins via Claude Code

Source: https://github.com/factory-ai/cursed-plugins/blob/main/README.md

Use the Claude Code plugin commands to add and install the toolkit.

```text
/plugin marketplace add Factory-AI/cursed-plugins
/plugin install cursed-plugins@cursed-plugins
```

--------------------------------

### Invoke /roast command for code review

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Use the /roast command to get a humorous, persona-driven review of your codebase. You will be prompted to select a persona and scope for the analysis.

```bash
# Invoke the roast command
/roast
```

--------------------------------

### Install Cursed Plugins via Git Clone

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Clone the Cursed Plugins repository and point your AI agent to the skills directory for manual integration. This method is suitable for agents other than Factory Droid or Claude Code.

```bash
# Clone the repository
git clone https://github.com/Factory-AI/cursed-plugins.git

# Point your agent at the skills directory
# Each skill is self-contained in skills/<name>/SKILL.md
```

--------------------------------

### Get Top 5 Contributors

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

Executes a git command to find the top 5 contributors by commit count. This is part of the initial discovery phase.

```bash
git log --format="%an" --no-merges -200 | sort | uniq -c | sort -rn | head -5
```

--------------------------------

### Construct X (Twitter) Share Link

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/cobol-converter/SKILL.md

This is an example of how to construct a Twitter share link using the pull quote and a hashtag. Ensure the text is URL-encoded and does not exceed character limits.

```javascript
https://x.com/intent/tweet?text=<URL-encoded tweet text>
```

--------------------------------

### Get Recent Commits with Stats

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

Fetches the last 20 commits with one-line summaries and short statistics about file changes. Useful for a quick scan of recent activity.

```bash
git log --oneline --shortstat --no-merges -20
```

--------------------------------

### Get Local Git User Name

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/obituary/SKILL.md

Retrieves the configured Git username for the local repository.

```bash
git config user.name
```

--------------------------------

### Get Midnight Coder Commits

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

Fetches the last 50 commits with detailed author, timestamp, and subject. Used to find suspicious commits made at odd hours for 'The Midnight Coder' focus.

```bash
git log --format="%H|%an|%ai|%s" --no-merges -50
```

--------------------------------

### Get All-Time Worst Commits

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

Retrieves the last 50 commits to identify the 'worst offender' based on commit size and message quality. Used for 'All-Time Worst' focus.

```bash
git log --oneline --shortstat --no-merges -50
```

--------------------------------

### Invoke /dating-profile command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Use the /dating-profile command to generate a dating app bio for your repository based on its technical characteristics. You will be prompted to select the scope of analysis.

```bash
# Invoke the dating profile command
/dating-profile
```

--------------------------------

### Clone Cursed Plugins Repository

Source: https://github.com/factory-ai/cursed-plugins/blob/main/README.md

Clone the repository to access the skills directory for manual integration.

```bash
git clone https://github.com/Factory-AI/cursed-plugins.git
```

--------------------------------

### Invoke Therapy Session Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Initiates the codebase counseling module to identify pain points and technical debt.

```bash
# Invoke the therapy session command
/therapy-session

# The module will prompt you to select:
# Scope: Whole repo / Specific folder or module / Specific contributor

# Looks for pain points:
```

--------------------------------

### Invoke /agent-repellent command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Use the /agent-repellent command to assess your codebase's resistance to AI coding agents. You will be prompted to select the scope of analysis.

```bash
# Invoke the agent repellent command
/agent-repellent
```

--------------------------------

### Execute Git Log Command

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/ship-ritual/SKILL.md

Use the Execute tool to run a git log command to display the last 10 commits in one-line format. This provides a concise history of recent changes.

```bash
git log --oneline -10
```

--------------------------------

### Invoke Feng Shui Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

This command initiates the architectural harmony assessment. It may prompt for scope selection.

```bash
/feng-shui
```

--------------------------------

### Execute Git Commands for Contributor Analysis

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/therapy-session/SKILL.md

Commands used to identify top contributors and the local user for scoping the therapy session.

```bash
git log --format='%an' --no-merges -200 | sort | uniq -c | sort -rn | head -5
```

```bash
git config user.name
```

--------------------------------

### Invoke COBOL Converter Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Runs the COBOL migration feasibility assessment tool to analyze repository architecture and provide migration estimates.

```bash
# Invoke the COBOL converter command
/cobol-converter

# The module will prompt you to select:
# Scope: Whole repo / Specific folder or module

# Scans for:
# - Lines of code by language
# - File counts and structure
# - Dependency counts from package.json/requirements.txt
# - Contributor count
# - Monorepo vs single app architecture

# Example output JSON schema:
{
  "type": "cobol-converter",
  "title": "COBOL Migration: my-project",
  "content": "Based on our assessment of 47,000 lines of TypeScript across 312 files, I recommend a 24-month migration timeline with a team of 8 senior COBOL developers. The estimated cost of $2.4M is a small price for mainframe reliability and a proven 60-year track record.",
  "pullQuote": "The estimated cost of $2.4M is a small price for mainframe reliability.",
  "repoName": "my-project"
}
```

--------------------------------

### Execute Git Diff Command

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/ship-ritual/SKILL.md

Use the Execute tool to run a git diff command to show the statistics of changes in the last 5 commits. This helps in identifying the scope of recent modifications.

```bash
git diff HEAD~5 --stat
```

--------------------------------

### Invoke Performance Review Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Executes the performance review module to analyze codebase patterns based on selected reviewer styles and scope.

```bash
# Invoke the performance review command
/performance-review

# The module will prompt you to select:
# 1. Style: Corporate HR / Disappointed Parent / Drill Sergeant / Therapist
# 2. Scope: Whole repo / Specific folder or module / Specific contributor

# Example output JSON schema:
{
  "type": "performance-review",
  "title": "Corporate HR Review: my-project",
  "content": "This review period has shown exciting potential for improvement in several key areas. The codebase demonstrates a consistent inability to commit to a single state management solution, which we view as a growth opportunity.",
  "pullQuote": "The codebase demonstrates a consistent inability to commit to a single state management solution.",
  "repoName": "my-project"
}
```

--------------------------------

### Invoke Ship Ritual Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

This command initiates the pre-deployment blessing process. It may prompt for scope selection.

```bash
/ship-ritual
```

--------------------------------

### Roast output JSON schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/roast/SKILL.md

Defines the structure for the final roast report output.

```json
{
  "type": "roast",
  "title": "<Persona Name> on <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Execute LOC Count Command

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/cobol-converter/SKILL.md

Use this command with Execute to count lines of code for specific file types. It's useful for gathering raw data for migration cost estimates.

```bash
find . -name '*.ts' -o -name '*.js' -o -name '*.py' | head -200 | xargs wc -l
```

--------------------------------

### Therapy Session Output Schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/therapy-session/SKILL.md

The required JSON structure for the therapy session report.

```json
{
  "type": "therapy-session",
  "title": "Session Notes: <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Agent Repellent Output Schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/agent-repellent/SKILL.md

The required JSON structure for the assessment report generated by the agent-repellent tool.

```json
{
  "type": "agent-repellent",
  "title": "Agent Repellent Assessment: <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Standard Output Schema for Modules

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

This JSON schema represents the standardized output for all Cursed Plugins modules, including type, title, content, pull quote, and repository name.

```json
{
  "type": "<module-name>",
  "title": "<Report Title>",
  "content": "<1-2 paragraph analysis>",
  "pullQuote": "<Most quotable line, max 200 chars>",
  "repoName": "<repository-name>"
}
```

--------------------------------

### Generate Blame Output JSON

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

The expected JSON output structure for the 'blame' plugin. Includes type, title, content, pullQuote, and repoName.

```json
{
  "type": "blame",
  "title": "Case File: <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Performance Review Output Schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/performance-review/SKILL.md

The required JSON structure for the performance review output.

```json
{
  "type": "performance-review",
  "title": "<Style> Review: <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Ship Ritual: Search for Risk Signals

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

This command searches changed files for specific risk signals like HACK, TODO, or FIXME comments.

```bash
grep -r "// HACK\|// TODO\|// FIXME" <changed-files>
```

--------------------------------

### X (Twitter) Share Link Format

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

This format outlines how to construct a Tweet text and URL for sharing module reports on X (Twitter), including the pull quote and hashtag.

```json
// Tweet text = pullQuote + " #cursedplugins" (max 240 chars total)
// URL: https://x.com/intent/tweet?text=<URL-encoded tweet text>
```

--------------------------------

### Grep for Merge Conflict Markers

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

Uses Grep to search for merge conflict markers within files. This indicates unresolved conflicts in the repository's history.

```bash
grep -E '<<<<<<<|=======|>>>>>>>'
```

--------------------------------

### Git contributor analysis command

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/roast/SKILL.md

Retrieves the top 5 contributors by commit count, excluding merges.

```bash
git log --format='%an' --no-merges -200 | sort | uniq -c | sort -rn | head -5
```

--------------------------------

### Output JSON Schema for COBOL Converter

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/cobol-converter/SKILL.md

The plugin generates a JSON object with a specific schema for the migration assessment. This includes the report type, title, content, a pull quote, and the repository name.

```json
{
  "type": "cobol-converter",
  "title": "COBOL Migration: <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Invoke Blame Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Executes the git history investigation tool to identify problematic commits and version control offenses.

```bash
# Invoke the blame command
/blame

# The module will prompt you to select:
# 1. Focus: Recent Commits / All-Time Worst / The Midnight Coder
# 2. Scope: Whole repo / Specific folder or module / Specific contributor

# Git commands used:
git log --oneline --shortstat --no-merges -20  # Recent commits
git log --oneline --shortstat --no-merges -50  # All-time worst
git log --format="%H|%an|%ai|%s" --no-merges -50  # Midnight coder

# Example output JSON schema:
{
  "type": "blame",
  "title": "Case File: my-project",
  "content": "The commit was made at 3:47 AM on a Sunday. The message read simply 'fix'. 847 files changed. No tests. The author's name appears 312 times in the blame history. Some cases solve themselves.",
  "pullQuote": "The commit was made at 3:47 AM on a Sunday. The message read simply 'fix'.",
  "repoName": "my-project"
}
```

--------------------------------

### Ship Ritual: Inspect Recent Changes

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

These commands are used by the ship ritual module to inspect recent changes in the repository, showing file statistics and commit logs.

```bash
git diff HEAD~5 --stat
```

```bash
git log --oneline -10
```

--------------------------------

### Grep for Meaningless Commit Messages

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/blame/SKILL.md

Uses Grep to search for common meaningless commit messages like 'fix', 'stuff', or 'wip'. This helps identify low-quality commit history.

```bash
grep -E 'fix|stuff|wip'
```

--------------------------------

### Dating Profile JSON Schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/dating-profile/SKILL.md

The required structure for the generated dating profile output.

```json
{
  "type": "dating-profile",
  "title": "<repo-name>'s Dating Profile",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Git contributor file history command

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/roast/SKILL.md

Lists the most recently modified files by a specific author.

```bash
git log --author="<name>" --name-only --no-merges -20
```

--------------------------------

### Search for Dead Code Signals

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/obituary/SKILL.md

Uses Grep to search for common indicators of dead or deprecated code within files. This includes keywords like '@deprecated', 'DEPRECATED', and markers like 'HACK', 'TEMP', or 'WORKAROUND'.

```bash
Grep to search for dead code signals within the chosen area: @deprecated, DEPRECATED, TODO.*remov, FIXME.*delet, commented-out function/class definitions, and HACK/TEMP/WORKAROUND markers.
```

--------------------------------

### Find Oldest Untouched Files

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/obituary/SKILL.md

Identifies files that have not been modified for the longest time. This can help pinpoint forgotten or abandoned code.

```bash
git log to find the oldest untouched files.
```

--------------------------------

### Invoke Obituary Command

Source: https://context7.com/factory-ai/cursed-plugins/llms.txt

Runs the dead code retrospective tool to identify and eulogize deprecated or unused code segments.

```bash
# Invoke the obituary command
/obituary

# The module will prompt you to select:
# Scope: Whole repo / Specific folder or module / Specific contributor

# Searches for dead code signals:
grep -r "@deprecated" .
grep -r "DEPRECATED" .
grep -r "TODO.*remov" .
grep -r "FIXME.*delet" .
grep -r "HACK\|TEMP\|WORKAROUND" .

# Example output JSON schema:
{
  "type": "obituary",
  "title": "In Memoriam: formatLegacyDate()",
  "content": "formatLegacyDate() was born on March 14, 2019, authored by a developer who has since left the company. It dreamed of parsing timestamps in a format that no longer exists. It is survived by moment.js, which it was meant to replace.",
  "pullQuote": "It dreamed of parsing timestamps in a format that no longer exists.",
  "repoName": "my-project"
}
```

--------------------------------

### JSON Output Schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/ship-ritual/SKILL.md

The expected JSON output schema for the ship-ritual plugin, including type, title, content, pullQuote, and repoName.

```json
{
  "type": "ship-ritual",
  "title": "Deployment Blessing: <repo-name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

--------------------------------

### Obituary JSON Output Schema

Source: https://github.com/factory-ai/cursed-plugins/blob/main/skills/obituary/SKILL.md

The expected JSON structure for the obituary report. It includes a title, content, a pull quote, and the repository name.

```json
{
  "type": "obituary",
  "title": "In Memoriam: <function or file name>",
  "content": "<plain text, no emojis, 1-2 short paragraphs>",
  "pullQuote": "<most quotable line from content, max 200 chars>",
  "repoName": "<repo-name>"
}
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.