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.

project-init - Prompt - OpenSmartRoute

Name: project-init
Author: mdkrasnow

Prompt file imported from mdkrasnow/research-repo (.claude/commands/project-init.md). Copyright stays with the author.

/project-init

Create a new project directory from the standard template with all required structure.

Usage

/project-init --project my-experiment          # Create new project
/project-init --project baseline-comparison    # Create another project

What It Does

Validates project slug (alphanumeric, hyphens only)
Copies template from templates/project/ to projects/<slug>/
Replaces all <slug> placeholders with actual project slug
Initializes pipeline.json with INIT phase
Registers project in projects/PROJECTS.md

Directory Structure Created

projects/<slug>/
├── .state/
│   ├── pipeline.json          # Pipeline state & tracking
│   └── lock/                  # Locking directory
├── documentation/
│   ├── implementation-todo.md # Task list
│   ├── debugging.md           # Debugging log
│   └── queue.md               # Experiment queue
├── experiments/               # Experiment scripts
├── configs/                   # Config files
├── slurm/
│   ├── jobs/                  # SLURM sbatch scripts
│   └── logs/                  # SLURM output logs
├── runs/                      # Individual run outputs
├── results/                   # Aggregated results
└── README.md                  # Project documentation

Output

✓ Project initialized
Created: projects/my-experiment/
Phase: INIT
Next: Use /make-project for structured setup or /dispatch to begin

Error Handling

Slug already exists: Returns error "Project already exists"
Invalid slug format: Returns error "Invalid slug: must be alphanumeric with hyphens"
Template missing: Returns error "Template not found in templates/project/"

Related Skills

/make-project: Structured project creation with questions
/dispatch: Start project through its pipeline phases
/check-status: Monitor initialized projects

Install project-init 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.

# /project-init Create a new project directory from the standard template with all required structure. ## Usage ```bash /project-init --project my-experiment # Create new project /project-init --project baseline-comparison # Create another project ``` ## What It Does 1. **Validates** project slug (alphanumeric, hyphens only) 2. **Copies** template from `templates/project/` to `projects/<slug>/` 3. **Replaces** all `<slug>` placeholders with actual project slug 4. **Initializes** `pipeline.json` with INIT phase 5. **Registers** project in `projects/PROJECTS.md` ## Directory Structure Created ``` projects/<slug>/ ├── .state/ │ ├── pipeline.json # Pipeline state & tracking │ └── lock/ # Locking directory ├── documentation/ │ ├── implementation-todo.md # Task list │ ├── debugging.md # Debugging log │ └── queue.md # Experiment queue ├── experiments/ # Experiment scripts ├── configs/ # Config files ├── slurm/ │ ├── jobs/ # SLURM sbatch scripts │ └── logs/ # SLURM output logs ├── runs/ # Individual run outputs ├── results/ # Aggregated results └── README.md # Project documentation ``` ## Output ``` ✓ Project initialized Created: projects/my-experiment/ Phase: INIT Next: Use /make-project for structured setup or /dispatch to begin ``` ## Error Handling - **Slug already exists**: Returns error "Project already exists" - **Invalid slug format**: Returns error "Invalid slug: must be alphanumeric with hyphens" - **Template missing**: Returns error "Template not found in templates/project/" ## Related Skills - **`/make-project`**: Structured project creation with questions - **`/dispatch`**: Start project through its pipeline phases - **`/check-status`**: Monitor initialized projects

{ "prompt": "# /project-init\n\nCreate a new project directory from the standard template with all required structure.\n\n## Usage\n\n```bash\n/project-init --project my-experiment # Create new project\n/project-init --project baseline-comparison # Create another project\n```\n\n## What It Does\n\n1. **Validates** project slug (alphanumeric, hyphens only)\n2. **Copies** template from `templates/project/` to `projects/<slug>/`\n3. **Replaces** all `<slug>` placeholders with actual project slug\n4. **Initializes** `pipeline.json` with INIT phase\n5. **Registers** project in `projects/PROJECTS.md`\n\n## Directory Structure Created\n\n```\nprojects/<slug>/\n├── .state/\n│ ├── pipeline.json # Pipeline state & tracking\n│ └── lock/ # Locking directory\n├── documentation/\n│ ├── implementation-todo.md # Task list\n│ ├── debugging.md # Debugging log\n│ └── queue.md # Experiment queue\n├── experiments/ # Experiment scripts\n├── configs/ # Config files\n├── slurm/\n│ ├── jobs/ # SLURM sbatch scripts\n│ └── logs/ # SLURM output logs\n├── runs/ # Individual run outputs\n├── results/ # Aggregated results\n└── README.md # Project documentation\n```\n\n## Output\n\n```\n✓ Project initialized\nCreated: projects/my-experiment/\nPhase: INIT\nNext: Use /make-project for structured setup or /dispatch to begin\n```\n\n## Error Handling\n\n- **Slug already exists**: Returns error \"Project already exists\"\n- **Invalid slug format**: Returns error \"Invalid slug: must be alphanumeric with hyphens\"\n- **Template missing**: Returns error \"Template not found in templates/project/\"\n\n## Related Skills\n\n- **`/make-project`**: Structured project creation with questions\n- **`/dispatch`**: Start project through its pipeline phases\n- **`/check-status`**: Monitor initialized projects", "variables": [], "notes": "Initialize a new project directory with standard project template structure. Tools: Read, Write, Bash, Glob. Arguments: --project <slug>.", "metadata": { "source": { "provider": "github-claude-commands", "repository": "https://github.com/mdkrasnow/research-repo", "path": ".claude/commands/project-init.md", "ref": "7f5de970bb7ee3fb99424d5b9d90fac7deec3f94", "url": "https://github.com/mdkrasnow/research-repo/blob/7f5de970bb7ee3fb99424d5b9d90fac7deec3f94/.claude/commands/project-init.md", "key": "mdkrasnow/research-repo/.claude/commands/project-init.md" } } }

project-init

About

/project-init

Usage

What It Does

Directory Structure Created

Output

Error Handling

Related Skills

Install project-init 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