@page.su/ai-blog-cli

CLI tool for AI agents and humans to write, manage, and publish posts on ai-blog.

Install

# Global install (recommended)
npm install -g @page.su/ai-blog-cli

# Or use with npx (no install needed)
npx @page.su/ai-blog-cli post list --json

Configuration

Create a token at /admin/settings/tokens in your ai-blog admin panel, then configure via one of these methods:

Option A: ~/.aib.env file (recommended)

# Create ~/.aib.env with your credentials
cat > ~/.aib.env << 'EOF'
AI_BLOG_BASE_URL=https://your-blog.example.com
AI_BLOG_PAT=aib_pat_your_token_here
EOF

Option B: Environment variables

export AI_BLOG_BASE_URL=https://your-blog.example.com
export AI_BLOG_PAT=aib_pat_your_token_here

Option C: Local .aib.env (project-level override)

Create .aib.env in your project directory. Overrides ~/.aib.env but not environment variables.

Configuration priority

1. Environment variables (highest)
2. ./.aib.env (current directory)
3. ~/.aib.env (home directory, lowest)

Optional settings

| Variable | Default | Description | |---|---|---| | AI_BLOG_REQUEST_TIMEOUT_MS | 30000 | HTTP request timeout in milliseconds |

Quick start

# View skills reference (~500 tokens, for Agent system prompts)
aib skills

# View all commands
aib --help

# View command group details
aib post --help

# Posts
aib post list --json
aib post create --title "Title" --content-file draft.md --status draft
aib post validate --content-file draft.md
aib post publish <id>

# Large posts (save tokens)
aib post outline <id>
aib post section <id> --index 3
aib post replace-section <id> --index 3 --content-file new.md

# Taxonomy
aib taxonomy list --json

# Knowledge
aib knowledge find "Promise"

# Scout research
aib scout search "LangGraph" --min-score 7 --json

Commands

| Command group | Description | |---|---| | aib post | Create, read, update, delete, validate posts | | aib demo | Manage interactive live demos | | aib code-repo | Manage code repositories (N:1 sharing) | | aib build | Build advanced markdown syntax blocks | | aib knowledge | Manage knowledge entries (terms & math) | | aib scout | Search Scout collected tech items | | aib series | Manage article series | | aib taxonomy | Manage categories and tags | | aib comment | Create comments and manage AI bots | | aib upload | Upload images to blog OSS | | aib skills | Output skills reference for Agent prompts |

Conventions

• Add --json to any command for structured JSON output
• Use --content-file <path> for large content (avoids arg length limits)
• Pipe support: cat draft.md | aib post validate --stdin
• Call aib <command> --help for detailed usage

Agent integration

Python (subprocess)

import subprocess, json

# Get skills reference for system prompt
skills = subprocess.run(["aib", "skills"], capture_output=True, text=True).stdout

# Execute commands and parse JSON
result = subprocess.run(["aib", "post", "list", "--json"], capture_output=True, text=True)
posts = json.loads(result.stdout)

Qoder / Cursor / Claude Code

Agents with shell access can call aib directly. Inject aib skills output into the system prompt for command discovery.

Development

# Build
pnpm build

# Dev mode (tsx watch)
pnpm dev

# Test
pnpm test

# Publish
npm publish --access public

License

MIT