QMD supports multiple output formats for both search results and document retrieval, designed for different use cases from human reading to machine parsing.

Available Formats

All formats are available for search, query, vsearch, and multi-get commands:

Flag	Format	Best For
`--json`	JSON	API integration, parsing, structured data
`--csv`	CSV	Spreadsheets, data analysis, tabular export
`--md`	Markdown	Documentation, reports, readable export
`--xml`	XML	Legacy systems, RSS feeds, structured exchange
`--files`	Files list	Quick reference, shell scripting
(default)	CLI	Human reading in terminal

JSON Format

Structured data with all metadata included. Usage:

qmd search "authentication" --json -n 3
qmd multi-get "docs/*.md" --json

Search Output Structure:

[
  {
    "docid": "#a3f2c1",
    "score": 0.84,
    "file": "qmd://docs/auth.md",
    "title": "Authentication Guide",
    "context": "Internal security documentation",
    "snippet": "...authentication flow uses JWT tokens..."
  }
]

Multi-Get Output Structure:

[
  {
    "file": "qmd://docs/readme.md",
    "title": "Project Overview",
    "context": "Main project documentation",
    "body": "# Project Overview\n\nThis is the main project..."
  },
  {
    "file": "qmd://logs/large-file.md",
    "title": "Large Log File",
    "skipped": true,
    "reason": "File too large (250KB > 10KB). Use 'qmd get qmd://logs/large-file.md' to retrieve."
  }
]

Key Fields:

docid: Short document ID (6-character hash prefix) - use with qmd get #abc123
score: Relevance score 0-1 (search only)
file: Virtual path in qmd://collection/path format
context: Folder context if configured (omitted if null)
snippet: Extracted relevant text (search) or body: Full content (multi-get)
skipped: Boolean indicating if file was too large (multi-get)

CSV Format

Tabular format for spreadsheet import and data analysis. Usage:

qmd search "error handling" --csv -n 10 > results.csv
qmd multi-get "*.md" --csv > all-docs.csv

Search Output:

docid,score,file,title,context,line,snippet
#a3f2c1,0.84,qmd://docs/errors.md,Error Handling,Engineering docs,42,"Error handling uses try/catch blocks..."
#b5d8e9,0.72,qmd://api/exceptions.md,Exception Types,API reference,18,"Custom exceptions inherit from BaseError..."

Multi-Get Output:

file,title,context,skipped,body
qmd://docs/readme.md,Overview,Main docs,false,"# Overview\n\nThis document..."
qmd://logs/big.md,Large File,,true,"File too large (50KB > 10KB)"

CSV Features:

Headers included in first row
Fields with commas/quotes/newlines are quoted per RFC 4180
Empty context fields show as blank (not “null”)
Compatible with Excel, Google Sheets, pandas

Markdown Format

Human-readable format suitable for documentation and reports. Usage:

qmd search "deployment" --md -n 5 > deployment-results.md
qmd multi-get "guides/*.md" --md > combined-guides.md

Search Output:

---
# Deployment Guide

**docid:** `#c7a4b2`
**context:** Production operations

Deployment happens through GitHub Actions which automatically builds
and deploys to staging environments...
---
# Container Configuration

**docid:** `#d9f1e3`
**context:** Production operations

Container images are built using multi-stage Dockerfiles to minimize
the final image size...

Multi-Get Output:

## qmd://guides/setup.md

**Title:** Initial Setup

**Context:** Getting started guides

Initial Setup

Follow these steps to get started:

Clone the repository
Install dependencies …

## qmd://guides/config.md

**Title:** Configuration

> File too large (45KB > 10KB). Use 'qmd get qmd://guides/config.md' to retrieve.

Markdown Features:

H1/H2 headings for document titles
Bold metadata labels
Code blocks for full content
Horizontal rules separate results (search)
Blockquotes for skip messages

XML Format

Structured format for legacy systems and RSS-like feeds. Usage:

qmd search "api" --xml -n 5 > api-docs.xml
qmd multi-get "changelog/*.md" --xml > changelog.xml

Search Output:

<file docid="#a3f2c1" name="qmd://api/rest.md" title="REST API" context="API documentation">
The REST API provides programmatic access to all system features.
Requests use JSON and return standard HTTP status codes...
</file>

<file docid="#b5d8e9" name="qmd://api/websocket.md" title="WebSocket API" context="API documentation">
WebSocket connections enable real-time bidirectional communication.
Connect to ws://api.example.com/v1/stream...
</file>

Multi-Get Output:

<?xml version="1.0" encoding="UTF-8"?>
<documents>
  <document>
    <file>qmd://docs/readme.md</file>
    <title>Project Overview</title>
    <context>Main documentation</context>
    <body># Project Overview

This project provides...</body>
  </document>
  <document>
    <file>qmd://logs/large.md</file>
    <title>Large Log</title>
    <skipped>true</skipped>
    <reason>File too large (80KB &gt; 10KB)</reason>
  </document>
</documents>

XML Features:

Self-closing document declaration
Attributes for metadata (search)
Elements for structured data (multi-get)
Proper XML entity escaping (<, >, &, ", ')

Files Format

Minimal format showing just paths, scores, and context. Usage:

qmd search "config" --files -n 20
qmd multi-get "*.md" --files

Search Output:

#a3f2c1,0.84,qmd://config/app.yml,"Application configuration files"
#b5d8e9,0.72,qmd://config/db.yml,"Database connection settings"
#c7a4b2,0.68,qmd://config/cache.yml,"Redis cache configuration"

Multi-Get Output:

qmd://docs/readme.md,"Main documentation"
qmd://docs/api.md,"API reference documentation"
qmd://logs/large.md,"Log files",[SKIPPED]

Format:

Search: #docid,score,path,"context"
Multi-get: path,"context" or path,"context",[SKIPPED]
Context quoted if present, omitted entirely if null
Ideal for piping to other tools: qmd search "term" --files | cut -d, -f3

CLI Format (Default)

Colored, human-readable output for terminal use. Usage:

qmd search "authentication"  # Uses CLI format by default
qmd multi-get "docs/*.md"

Search Output:

qmd://docs/auth.md  #a3f2c1  [0.84]
Context: Internal security documentation
────────────────────────────────────────────────
...authentication flow uses JWT tokens signed
with RS256. Tokens expire after 1 hour and must
be refreshed using the /auth/refresh endpoint...

qmd://api/oauth.md  #b5d8e9  [0.72]
Context: API reference documentation
────────────────────────────────────────────────
OAuth2 implementation follows RFC 6749 with
support for authorization code and client
credentials grant types...

Multi-Get Output:

============================================================
File: qmd://docs/readme.md
============================================================

Folder Context: Main project documentation
---

# Project Overview

This is the main project documentation...

============================================================
File: qmd://logs/large.md
============================================================

[SKIPPED: File too large (80KB > 10KB). Use 'qmd get qmd://logs/large.md' to retrieve.]

CLI Features:

Syntax highlighting and colors (disabled if NO_COLOR env var set)
Visual separators between results
Compact score display
Context shown when available
Optimized for terminal width
Skip messages for large files

Format-Specific Options

Line Numbers

Add line numbers to output (all formats except —files):

qmd search "error" --json --line-numbers
qmd get readme.md --line-numbers

Output (JSON example):

{
  "snippet": "1: # Error Handling\n2: \n3: All errors inherit from BaseError..."
}

Full Content

Show complete documents instead of snippets (search only):

qmd search "deploy" --full --md
qmd query "authentication flow" --full --json

Without --full: Returns ~300 character snippets around matching text With --full: Returns entire document content

Default Limits by Format

Format	Default -n	Reason
CLI	5	Terminal readability
JSON	20	API pagination
CSV	5	Focused results
Markdown	5	Document length
XML	5	Payload size
Files	20	Compact listing

Override with -n <number> or --all (returns up to 100,000 results with --min-score filtering).

Common Patterns

Export to file

qmd search "security" --json > security-results.json
qmd multi-get "*.md" --csv > all-docs.csv

Pipe to processing tools

qmd search "error" --files | cut -d, -f3 | xargs qmd get
qmd search "deploy" --json | jq '.[] | select(.score > 0.8)'

Combine with grep/awk

qmd search "api" --csv | grep "authentication"
qmd multi-get "*.md" --files | awk -F, '{print $1}' 

Import to spreadsheet

qmd search "performance" --csv -n 50 > perf.csv
# Open perf.csv in Excel/Google Sheets

Generate documentation

qmd multi-get "docs/**/*.md" --md > combined-docs.md
qmd search "changelog" --md --full > release-notes.md

Multi-Get Size Limits

By default, multi-get skips files larger than 10KB (10,240 bytes) to prevent overwhelming output. Adjust the limit:

qmd multi-get "*.md" --max-bytes 50000  # 50KB limit
qmd multi-get "*.md" --max-bytes 1000000 --json  # 1MB limit

Skipped file output:

JSON: {"file": "...", "skipped": true, "reason": "File too large..."}
CSV: "...",true,"File too large..."
Markdown: > File too large (80KB > 10KB). Use 'qmd get ...' to retrieve.
XML: <skipped>true</skipped><reason>...</reason>
Files: ...,[SKIPPED]
CLI: [SKIPPED: File too large...]

Retrieve skipped files:

qmd get qmd://large-file.md  # No size limit on single get

Commands

Options

Output Formats

Available Formats

JSON Format

CSV Format

Markdown Format

Initial Setup

XML Format

Files Format

CLI Format (Default)

Format-Specific Options

Line Numbers

Full Content

Default Limits by Format

Common Patterns

Export to file

Pipe to processing tools

Combine with grep/awk

Import to spreadsheet

Generate documentation

Multi-Get Size Limits

See Also

Build docs developers (and LLMs) love

Commands

Options

Documentation Index

​Available Formats

​JSON Format

​CSV Format

​Markdown Format

​Initial Setup

​XML Format

​Files Format

​CLI Format (Default)

​Format-Specific Options

​Line Numbers

​Full Content

​Default Limits by Format

​Common Patterns

​Export to file

​Pipe to processing tools

​Combine with grep/awk

​Import to spreadsheet

​Generate documentation

​Multi-Get Size Limits

​See Also

Build docs developers (and LLMs) love

Available Formats

JSON Format

CSV Format

Markdown Format

Initial Setup

XML Format

Files Format

CLI Format (Default)

Format-Specific Options

Line Numbers

Full Content

Default Limits by Format

Common Patterns

Export to file

Pipe to processing tools

Combine with grep/awk

Import to spreadsheet

Generate documentation

Multi-Get Size Limits

See Also