@destiner/pi-subagents
Subagents for pi: delegate by intelligence level, and let your settings decide which model runs it
$ pi install npm:@destiner/pi-subagentsPackage Catalog
Extensions, skills, prompt templates, and themes published to npm. Install with pi install npm:<package>. See the package docs for details.
Recently published
@mu-crew/muA persistent, observable crew of coding agents (pi, Claude Code, Codex, cursor-agent, Muse Code) in tmux, herdr or Rex panes, coordinated through a built-in task DAG.11m agopi-cache-optimizerImprove Pi prompt/KV cache hit rates with stable prompts, OpenAI-compatible cache keys, proxy compat warnings, and footer cache stats.18m agokeyshroudLet AI coding agents use your secrets without ever seeing them. Encrypted local vault or self-hosted Vaultwarden/Bitwarden, masked output, MCP server, Claude Code + pi plugins.19m agoprojectopsOne npx install for GitHub release automation: semantic versioning, changelogs, release PRs, CI/CD workflows and AI agent skills19m agopi-straw-bossStraw Boss coordination for pi in Herdr: sessions join a tab's net by choice, the first joiner is its main agent, a workspace tower opens on demand with a live monitoring dashboard, and boss_broadcast reaches the rest20m ago@cortexkit/pi-anthropic-authUse Claude Pro/Max subscriptions in [Pi](https://pi.dev) with `@cortexkit/pi-anthropic-auth`. The companion `@cortexkit/opencode-anthropic-auth` package supports [OpenCode](https://opencode.ai) 1 and 2; both integrations use the shared `@cortexkit/anthrop24m agopi-optchatOptChat memory for Pi: persistent profiles, zoomable history, background agents, and conversation imports.30m ago@sassywares/skillsLean, explicit-invocation Agent Skills for shipping fast: plan, work, review, validate, ship.32m ago
All packages
4051-4100 / 5387
@normful/pi-auto-name
Auto-generates names in your language for Pi sessions, tmux windows, herder panes/tabs, zellij panes/tabs so you can easily view what each Pi is doing
$ pi install npm:@normful/pi-auto-namepi-bedrock-mantle
Pi extension: all Bedrock Mantle models (GPT-5.x, Claude, DeepSeek, Qwen3, Mistral, Kimi, ...) via SigV4 — no long-term API key needed
$ pi install npm:pi-bedrock-mantlepi-better-auto-compact
pi coding agent extension: additional auto-compact thresholds (percent of context window / used tokens), taking the lowest of all thresholds including pi's built-in remaining-tokens threshold
$ pi install npm:pi-better-auto-compactpi-cache-stats
Pi Agent extension for prompt-cache observation and prefix diagnostics
$ pi install npm:pi-cache-statspi-model-dock
pi-model-dock: an interactive custom provider manager for Pi and Oh My Pi.
$ pi install npm:pi-model-dock@rabden/pi-opencode-zen
Pi extension: OpenCode Zen free models (Responses endpoint) with stream-translated tool calling
$ pi install npm:@rabden/pi-opencode-zen@aeondave/pi-skill-optimizer
Less repeated context for Pi: local skill routing, on-demand tools, and recoverable output reduction.
$ pi install npm:@aeondave/pi-skill-optimizer@esuyo/pi-esuyo-custom-provider
Register custom OpenAI-compatible providers in Pi.dev via JSON config
$ pi install npm:@esuyo/pi-esuyo-custom-provider@hk_net/remote-pi
Mobile remote control and local agent mesh for the Pi coding agent. Pair your phone via QR over a relay, watch tool calls in real time, run multi-Pi sessions over a local Unix Domain Socket broker, and let agents talk to each other through structured requ
$ pi install npm:@hk_net/remote-pipi-flow-memory-gate
Memory gate for Pi: a local decision model decides what is worth remembering, you approve. Additive to pi-hermes-memory.
$ pi install npm:pi-flow-memory-gatepi-ws-recovery
WebSocket→SSE fallback indicator and automatic WebSocket recovery for Pi's OpenAI Codex provider
$ pi install npm:pi-ws-recoverypi-zed-bridge
Pi extension for Zed diff approval workflows and editor context bridging.
$ pi install npm:pi-zed-bridge@ayagmar/summarize
Fetch a URL or convert a local file (PDF, DOCX, HTML, and more) into Markdown with markitdown, optionally summarizing it.
$ pi install npm:@ayagmar/summarize@fyeeme/pi-mermaid-viewer
Render Mermaid diagrams from the conversation in the browser.
$ pi install npm:@fyeeme/pi-mermaid-viewerpi-autocontext
autocontext extension for Pi coding agent — iterative strategy generation, LLM judging, and evaluation tools
$ pi install npm:pi-autocontextpi-cursor-bridge
Windows only. Use Cursor Context Engine and bounded, explicitly continuous Cursor Agent execution from the Pi coding agent.
$ pi install npm:pi-cursor-bridgepi-response-timer
Live elapsed time and a minimal completion timestamp for Pi's terminal UI.
$ pi install npm:pi-response-timerpi-closeout
Session closeout gate for coding agents: flags staged-but-uncommitted sweep fuel, untracked stash-bait, dirty child repos, and open task claims when a session ends — plus Work-Session identity rows.
$ pi install npm:pi-closeoutpi-image-compact
Pi extensions that keep screenshot-heavy sessions small: near-lossless WebP, a request image budget, pruning after compaction, and a footer indicator.
$ pi install npm:pi-image-compactpi-provider-alibaba
Alibaba Cloud DashScope provider for pi (coding, token, and pay-as-you-go, both international and Chinese endpoints)
$ pi install npm:pi-provider-alibabapi-qoder-agent-sdk-bridge
Pi extension that bridges the Qoder Agent SDK (official @qoder-ai/qoder-agent-sdk) into Pi as a model provider
$ pi install npm:pi-qoder-agent-sdk-bridge@sidshaytay/pi-llama-swap
Pi extension that discovers llama-swap models and registers them as a provider.
$ pi install npm:@sidshaytay/pi-llama-swap@brettinternet/pi-progress
Compact passive activity and verification progress for Pi
$ pi install npm:@brettinternet/pi-progress@liuchen64/pi-im-bridge
Pi 扩展 - IM 桥接，通过即时通讯工具远程与当前 pi 会话交互（内置微信通道）
$ pi install npm:@liuchen64/pi-im-bridgepi-context-guard-jev
Pi extension that removes tool output the agent no longer needs from its context. The Jev classifier keeps the needed lines verbatim; stale small outputs, unrelated old exchanges and old screenshots become one-line notes. Everything can be recalled.
$ pi install npm:pi-context-guard-jevpi-unstall
Pi extension that notices when an agent run stopped without finishing - provider error, truncated response, or a model that trailed off mid-sentence - and nudges it back into motion with a continue message.
$ pi install npm:pi-unstall@ayagmar/release-readiness-certification
Produce a strict evidence-based release verdict with severity-ranked findings, blockers, a pre-release work batch, and a deferred backlog.
$ pi install npm:@ayagmar/release-readiness-certification@iantri/pi-sandbox-aware
Make pi aware when it is running inside a macOS sandbox
$ pi install npm:@iantri/pi-sandbox-awarepi-provider-gdpr-inference
Pi provider for Entrim — a GDPR-compliant EU inference platform (OpenAI-compatible, live model discovery with real per-model pricing)
$ pi install npm:pi-provider-gdpr-inferenceprime-cursor-sdk
Prime Agent provider extension backed by the Cursor SDK local and cloud agents
$ pi install npm:prime-cursor-sdk@imolko/ultra-agent
Metodología ultra como paquete versionado: skills canónicos, registro ADR A/T/P, extensiones de pi y CLI (install/update/status/ship).
$ pi install npm:@imolko/ultra-agentjackwu-pck-test
Pi package with Chinese staged-change review prompts and a Conventional Commits message skill.
$ pi install npm:jackwu-pck-testpi-minimal-claude
Pi package with planner/worker/explorer/debugger agents and manual|code-only|auto approval modes
$ pi install npm:pi-minimal-claudepi-provider-status
Show remaining quota and balance for LLM providers in pi — /status panel for Kimi Coding, Zhipu ZAI, MiniMax, DeepSeek, OpenRouter, Claude and Codex subscription windows
$ pi install npm:pi-provider-statuspi-readonly-mode
Strict read-only mode extension for pi: blocks edit/write and destructive bash commands, toggled with /readonly or --readonly.
$ pi install npm:pi-readonly-mode@tifan/pi-mermaid-open
Find skipped Mermaid diagrams and show them in a terminal image viewer.
$ pi install npm:@tifan/pi-mermaid-open@ayagmar/advisor-executor-quality-loop
Orchestrate complex multi-task work by separating advisor reasoning, delegated execution, and independent verification until tasks converge.
$ pi install npm:@ayagmar/advisor-executor-quality-looppi-friends
pi extension: consult other models on hard problems. Each friend (any OpenAI-compatible endpoint) gets its own ask_<name> tool; briefs are self-contained one-turn exchanges, answers come back without reasoning transcripts.
$ pi install npm:pi-friendspi-local-models
Pi extension for any local model harness - model discovery, per-project config
$ pi install npm:pi-local-models