How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

shell-is-all-agents-need - Skill

Imported from kyuheon-kr/shell-is-all-agents-need (AGENTS.md). Install upstream with npx skills add kyuheon-kr/shell-is-all-agents-need. Copyright stays with the author.

Working agreements

Develop on a task branch and finish requested changes by committing, pushing that branch, and opening a pull request against main. Never push changes directly to main. Do not ask for routine confirmation; follow explicit requests to hold a push or leave changes uncommitted.
Merge only when requested, through a pull request with required CI passing. Never bypass branch protection. Releases and npm publishing are separate, explicitly requested operations; follow docs/releasing.md.
Stage only files belonging to the task. Preserve unrelated user changes; never discard them or force-push without explicit authorization.
Report the resulting branch, commit hash, push status, and relevant validation results when finishing.
Keep the README compact. Prioritize quick setup, major use cases, and essential commands; avoid duplicating tool descriptions or lengthy implementation details.
Prefer the shell-is-all-agents-need MCP tools for shared-terminal work and terminal-based agent orchestration. Inspect the current program before sending input, use explicit pane IDs, and leave unrelated panes alone.
When launching worker agents, specify the repository working directory explicitly, assign clear tasks and completion markers, and use separate worktrees for concurrent edits. Read results and send follow-ups explicitly; subscriptions do not automatically instruct or wake agents.
For code changes, run npm run check and npm run build. Run npm run test:integration when changing Zellij commands, process handling, or subscriptions. For documentation-only changes, review the diff and run git diff --check.
Diagnose CLI failures using bounded output and command context. Retry only understood transient failures in safe read operations; never blindly replay mutations or suppress malformed output.

Install shell-is-all-agents-need and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

shell-is-all-agents-need

About

Working agreements

Install shell-is-all-agents-need and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Use it

Manifest

Reviews