Official agent plugin published by Tsuga, shipped as a single marketplace.
Operating Tsuga itself: the tsuga CLI driver (TQL, aggregations, deep links), live-platform investigation (service health, errors, latency, monitor coverage), dashboard building, an incident-investigation orchestrator, meta-skills for building knowledge-company / incident-history bundles, and OpenTelemetry instrumentation guidance across nine languages (Python, Go, Node.js, Java, .NET, Ruby, Rust, PHP, C++): SDK setup, traces, metrics, logs, trace-log correlation, Collector configuration, OTTL, semantic conventions, instrumentation-quality audits, and telemetry debugging. Live Tsuga workflows require a Tsuga account.
The Tsuga CLI adds the marketplace and installs the plugin in one command:
npm install -g @tsuga/cli
# Claude Code
tsuga install plugin claude-code
# Codex
tsuga install plugin codexFor Claude Code, this enables marketplace auto-update so the plugin refreshes on its own (pass --no-auto-update to skip).
Or add the marketplace and install the plugins directly:
# Claude Code
claude plugin marketplace add tsuga-dev/agent-plugins
claude plugin install tsuga@tsuga
# Codex
codex plugin marketplace add tsuga-dev/agent-plugins
codex plugin add tsuga@tsugaTsuga owns and maintains this plugin — anything generic about operating Tsuga or instrumenting with OpenTelemetry belongs here, and installs with autoUpdate receive new versions automatically.
- Found a gap or an error? Open an issue or PR. Field-tested corrections (a workflow the skill should cover, a gotcha it gets wrong) are exactly what we want flowing back upstream.
- Org-specific conventions (your naming schemes, internal runbooks, metric families, team structure) don't belong in these skills. Keep them in your own plugin layered on top — skills compose, and yours can reference these by name (e.g.
tsuga-cli).