INT
Apr 26

Just Bash Skills Agent

Minimal ToolLoopAgent chat demo: on-call-style skills (incident triage + stakeholder comms) on disk, skill tool loads SKILL.md + paths, readFile pulls reference checklists and templates from the sandbox.

View Full Pattern on Desktop
Just Bash Skills Agent preview

Problems Solved

  • Filesystem skill packs with readFile-only sandbox access
  • Minimal pattern for skill tool body vs reference files on disk
  • Streaming chat UI using AI Elements

Use Cases

  • Teaching skill loading and reference reads
  • Starter for internal runbook / ops copilots
  • Sandboxed readFile for bundled markdown references

AI SDK APIs

ToolLoopAgentcreateAgentUIStreamResponsegatewaystepCountIs

Tags

aiagentsbash-toolskillsreadfileagent-patternai-sdk

Requirements

Node.js
24+
Memory
512MB
Other
  • Vercel AI Gateway or provider API access
  • bash-tool runtime

Environment Variables

External Services

anthropic

Ready to use?

View the full interactive pattern with live preview and copy the source code.

View Pattern on Desktop

Related Patterns

Sub-Agent Orchestrator preview

Sub-Agent Orchestrator

Custom Agent implementation demonstrating the Agent interface abstraction with an orchestrator that routes queries to specialized sub-agents (research, analysis, support). Shows options passing and structured outputs.

Marketing agent with skills + readFile preview

Marketing agent with skills + readFile

Chat UI wired to a ToolLoopAgent that loads filesystem skills and exposes readFile on the skill sandbox—demo includes ad-creative, copywriting, and paid-ads skill packs for campaign workflows.

Research Agent Chain preview

Research Agent Chain

Sequential three-agent chain demonstrating structured outputs flowing between agents. Research Agent 1 gathers research, Expand Agent 2 expands research, Synthesis Agent 3 synthesizes final answer. Uses Exa tools and AI SDK 6's stabilized structured output support.

Marketing Plan & Implement Agent preview

Marketing Plan & Implement Agent

Two-phase marketing agent with brand context: Stage 0 extracts brand info from URL or manual input, planning phase gathers requirements and creates a plan with todos, implementation phase executes the plan using web scraping and research tools. Uses callOptionsSchema + prepareCall for dynamic agent configuration.

End of Pattern