Agent-Skills.md

Agent Skills: agent-friendly-cli

Use when writing command-line scripts agents will control

UncategorizedID: sanand0/scripts/agent-friendly-cli

Repository

sanand0/scripts
sanand0
11

Install this agent skill to your local

pnpm dlx add-skill https://github.com/sanand0/scripts/tree/HEAD/agents/agent-friendly-cli

Skill Files

Browse the full folder contents for agent-friendly-cli.

Download Skill

Loading file tree…

agents/agent-friendly-cli/SKILL.md

Skill Metadata

Name
agent-friendly-cli
Description
Use when writing command-line scripts agents will control
  • Support JSON input via --json or --params and output via --output json or --format json. Prefer structured output as default for non-TTY.
  • Validate all inputs aggressively. Humans and agents make mistakes or can be adversarial. Support --dry-run for writes. Confirm before destructive actions
  • Add a schema or --describe capability that returns machine-readable method signatures: params, request body, response types, ...
  • Fail fast showing reason and correct invocation.
  • Design for context-window efficiency. Document and support filters, field masks, NDJSON pagination, etc. so agents can request only needed data
  • Support environment-variable-driven, headless execution for unattended use (tokens, credential file paths)
  • Log state/progress BEFORE actions and flush - so agents can monitor, write errors that help agents debug, warn agents of unexpected input, data, configs, etc.
  • Document with compact, explicit instructions and examples
agent-friendly-cli Skill | Agent Skills