Find your next open-source project

Search by name, purpose, or keyword — or paste a GitHub repository URL
All skills

diegosouzapw/OmniRoute

Gives coding agents a shared set of commands and workflows for running and managing the OmniRoute AI gateway.

  • 52 skills
  • 85k stars

The 5 most installed

  • Sends tasks to the OmniRoute A2A server and inspects agent-to-agent skill history from the command line.

  • Triggers backups, syncs snapshots to cloud storage, and restores OmniRoute data from the command line.

  • Submits and monitors batch inference jobs, uploads files, and retrieves results from the command line.

  • Sends chat completions, streams responses, and starts an interactive chat session from the command line.

  • Configures and tests prompt compression rules and previews compressed output from the command line.

The other 47 skills
  • Manages context configurations, filter sets, and conversation sessions from the command line.

  • Shows cost breakdowns, token usage, and call logs, and exports usage reports from the command line.

  • Creates and runs evaluation suites, tracks benchmark progress, and compares model scores from the command line.

  • Checks server health, component status, and live metrics with a real-time dashboard from the command line.

  • Creates, lists, rotates, and revokes API keys and manages provider OAuth flows from the command line.

  • Inspects MCP server status, lists tools, runs tool calls, and manages audit logs from the command line.

  • Queries available AI models, lists aliases, and searches the model catalog from the command line.

  • Lists, installs, tests, and removes skills and plugins, and manages persistent memory from the command line.

  • Inspects audit logs, access policies, telemetry, and request history for compliance from the command line.

  • Lists, adds, tests, and validates provider connections and rotates their API keys from the command line.

  • Inspects circuit breakers, cooldowns, and quota limits, and resets stuck providers from the command line.

  • Creates, tests, and deletes routing combos and configures fallback chains from the command line.

  • Starts, stops, and restarts the OmniRoute server, managing daemon mode and ports from the command line.

  • Runs initial setup, configures global settings and environment variables, and checks for updates.

  • Detects installed coding tools, finds matching agent skills on GitHub, and installs them into those tools.

  • Starts and stops tunnels, shows active tunnel URLs, and tests external reachability from the command line.

  • Configures the OpenAI Codex CLI on Linux, macOS, or Windows to use OmniRoute as its backend.

  • Checks a queue of pull requests against their base, fixes authorized defects, and reports validation results.

  • Prepares and merges an owner-selected batch of pull requests with contributor credit and validation.

  • Implements authorized fixes, validates exact candidates, and merges only owner-approved pull request groups.

  • Lets agents talk to OmniRoute over JSON-RPC to route requests, check quotas, find providers and report costs.

  • Creates, lists, rotates and revokes API keys with per-key scopes, spending limits and expiry dates.

  • Sets up Bearer token and session authentication so requests to the OmniRoute API are accepted.

  • Sets spending limits, token quotas and rate limits per key or globally, and shows current consumption.

  • Shows cache statistics, lists and clears cached LLM responses, and manages the reasoning replay cache.

  • Lists, configures and invokes CLI tool plugins that extend what OmniRoute can automate.

  • Builds routing combos with priority, weighted or round-robin strategies, sets fallback chains and checks results.

  • Configures RTK and Caveman compression modes and rules to cut prompt tokens by 60–90%.

  • Sets up RTK filters and context rules, then tests compression on real prompt samples.

  • Lists and exports database backups, and reports system version, metrics and environment repair status.

  • Finds agent skills on GitHub, scores and scans them for secrets or malware, then installs them.

  • Calls OpenAI-compatible endpoints for chat, embeddings, images, audio, moderation, rerank and responses.

  • Connects to the OmniRoute MCP server over SSE, stdio or HTTP to use routing, cache and memory tools.

  • Lists available AI models across configured providers, resolves aliases and browses the full catalog.

  • Adds, updates, removes and tests AI provider connections, including API keys and OAuth flows.

  • Sets HTTP, HTTPS or SOCKS proxies per provider or globally, tests connectivity and manages rotation.

  • Monitors provider health, circuit breakers, latency percentiles, cooldowns and budget alerts in real time.

  • Reads and updates global settings such as system prompts, thinking budget, IP filters and payload rules.

  • Syncs OmniRoute configuration, provider connections and settings to and from cloud storage.

  • Creates ngrok, Cloudflare or custom tunnels to expose OmniRoute to remote agents and CI pipelines.

  • Filters call logs by provider, model, time, status and cost, then exports and aggregates token usage.

  • Installs, starts, stops, restarts and updates embedded services, and shows their status and logs.

  • Registers, tests and removes webhook endpoints, subscribes to events and manages delivery retries.

  • Pushes the agent toward the simplest working solution, reusing existing code and standard libraries first.

  • Runs a read-only quality scan and reports exact findings, failures and frozen debt before a release.

  • Fixes confirmed CI failures on the base branch first, then prepares each affected pull request with exact evidence.

  • Runs the local Release-Green check on a base or combined pull request candidate and separates hard failures from advisory drift.