Standalone home for '/run-scenario' (drive a Claude agent through a scenario file in tmux) and '/sync-jsonl-projects' (back up Claude Code session JSONL files and file-history), usable from any directory via --plugin-dir.
Find available domains and brandable names with instant.ai. Check names across extensions, verify domains in bulk, generate alternatives, and search USPTO trademark records. Free, with no account or API key required. Instant may earn referral fees on registrations completed on instantdomainsearch.com.
Runs a per-capability spec → test → impl → incident flow: humans approve and freeze traceable tests before any implementation, AI generates disposable impl to satisfy them, and every failure is wired to a regression test. Provides the skill plus /aiable-spec, /aiable-test, /aiable-impl, /aiable-incident commands.
[DEV CHANNEL — Medley dev team only] Prerelease Medley builds, used to test engine and daemon upgrades before they reach the stable channel. Expect breakage. Install this INSTEAD of the stable medley plugin, never alongside it: the two share ~/.medley, port 8730, and one background daemon. Missions inside Claude Code: give /mission a goal and Medley plans the work, then supervises execution until the goal is met.
Realtime inference API for open-source models in your agent: DeepSeek V4.1 Flash and Gemma 4 31B, plus Qwen 3 TTS and Nemotron ASR. Test the models on your own prompts, add a key to your code, speak text and transcribe audio.
Merget is a semantic merge queue for GitHub. This plugin connects your coding agent to Merget's MCP server and adds four skills: set Merget up, operate its merge queue, read a pull request's findings and verdict, and query its code property graphs.
Goal in, working code out: iterative convergence with verification loop and AI behavioral enforcement. Includes plan-ops.sh for plan lifecycle management. Four-phase workflow: Plan -> Review -> Implement -> Verify.
Kilo/Claude ecosystem meta-tooling - manage and author Kilo Code and Claude Code's own configuration: maintaining Kilo's RAG index, and generating a grounded reference skillset for a framework/library from its source/docs/spec via /generate-skillset. Not about delegating work - see ai-architect-executor/kilo-mcp for that.
Retrieve, inject, and manage secrets from Keeper Vault using KSM CLI (ksm) for developer workflows, including ksm exec, interpolate, and Keeper notation.
Iteratively evaluate and enhance plugins with Anthropic best practices. Analyzes quality, suggests concrete improvements, and integrates with Ralph Loop for continuous self-improvement.
Offline, citable Apple developer documentation search with the cupertino CLI (verified against cupertino v1.5.0): docs, App Review Guidelines, HIG, sample code, Swift Evolution, WWDC transcripts, Swift packages, symbol and inheritance queries, per-platform version filters, and SDK freshness checks.
Router for visionOS / Apple Vision Pro spatial computing work: picks the right skill from Tom Krikorian's visionOSAgents set (spatial SwiftUI, RealityKit, ARKit providers, ShaderGraph and USD authoring, SharePlay, WidgetKit, immersive media, Swift Charts 3D). Install those skills from tomkrikorian/visionOSAgents.