cited 0.1.0

CLI for running cited AI agents — one command, reproducible results

cited

Python CLI for running science-backed AI agents. Discover peer-reviewed agents and run them with a single command.

Installation

pip install cited

Quick Start

# Find an agent
cited search "failure mode analysis"

# See what it expects
cited info dfmea-agent

# Run it
cited run dfmea-agent --input input.json

# Validate input before running
cited validate dfmea-agent --input input.json

Commands

Command	Description
cited run <agent>	Run an agent with JSON input
cited search <query>	Search the agent catalog
cited info <agent>	Show agent details and schema
cited validate <agent>	Validate input without running
cited cache list	List cached agents
cited cache clear	Clear the local cache
cited config show	Show current configuration
cited config set <key> <value>	Update a config value
cited version	Show CLI version

Agent references

Agents can be referenced by slug or with a pinned version:

cited run dfmea-agent              # latest
cited run dfmea-agent@1.0.0       # pinned version

Input / output

Input is read from a file (--input) or stdin:

cited run dfmea-agent --input input.json
cat input.json | cited run dfmea-agent

Output is JSON written to stdout, making it easy to pipe:

cited run dfmea-agent --input input.json | jq '.results'

Configuration

Env var	Default	Description
CITED_API_URL	https://citedagents.com	Platform API base URL
CITED_CACHE_DIR	OS cache dir	Where agent blueprints are cached

Or use cited config set:

cited config set api-url https://citedagents.com
cited config set cache-dir /tmp/cited-cache

License

MIT