Plugin registry

Plugins for pansy, OpenCode and Pi, ranked by what npm, GitHub and their users say.

  1. 2701
    W

    Temporary package placeholder for staged publishing

  2. 2702
    P

    Automatically name Pi sessions from conversation context

  3. 2703
    P

    Context-saving MCP extension for Pi Coding Agent with sandboxed code execution, FTS5 knowledge base, and intent-driven search.

  4. 2704
    P

    Google Calendar, Gmail, Tasks, Drive, Contacts, Forms, Docs & Sheets MCP server for Pi — 36 tools, login once, manage Google from your AI agent.

  5. 2705
    S

    SPARK skills and runtime bootstrap for coding agents

  6. 2706
    P

    Pi-native ralph loop — autonomous coding iterations with mid-turn supervision

  7. 2707
    P

    Run bounded, read-only Pi subagents without filling the parent context with intermediate investigation output

  8. 2708
    I

    Respan instrumentation plugin and pi extension for the pi coding agent (@earendil-works/pi-coding-agent)

  9. 2709
    P

    Pi extension for Azure AI Foundry — discovers deployments and registers models

  10. 2710
    P

    A one-line recap of where the session stands when you come back to Pi, plus /recap

  11. 2711
    P

    Command Code Provider API for Pi — every top model (Claude, GPT, Gemini, DeepSeek, Kimi, GLM, Qwen, Grok & more) behind one OpenAI- or Anthropic-compatible endpoint, with the live model catalog discovered automatically at startup.

  12. 2712
    P

    After a task, flags the one thing you might have missed

  13. 2713
    P

    Voice in + voice out for Pi CLI — local-only hold-to-talk STT for Chinese and English (21 offline models) plus offline TTS (Kitten Nano, Piper, Kokoro); no audio is uploaded to a remote speech service

  14. 2714
    P

    Adds /goal to pi: conversational goal planning, flexible or ordered goals, persistent progress, and an independent completion auditor.

  15. 2715
    P

    Enforce structured design→execute→finalize workflow with TDD discipline in AI coding agents

  16. 2716
    P

    Web search (Exa + Brave) and content extraction for Pi

  17. 2717
    P

    Deep research skill for pi — structured search, reflection, and analysis.

  18. 2718
    P

    Deny-first typo-path model judge — a pi-permission-system Authorizer chain link

  19. 2719
    P

    Bounded UTF-8 child-process execution for Pi packages.

  20. 2720
    C

    Tells you when your Pi session has reached a good point to compact: a standalone extension that judges completed checkpoints instead of leaving you to guess.

  21. 2721
    P

    Use OpenRouter presets (@preset/<slug>) as first-class models in Pi, with live provider pricing in /presets.

  22. 2722
    R

    Continue a coding session in a different agent, or a different profile.

  23. 2723
    U

    Metodología ultra como paquete versionado: skills canónicos, registro ADR A/T/P, extensiones de pi y CLI (install/update/status/ship).

  24. 2724
    C

    Pi skill that calculates CVSS v4.0 scores and severity ratings, and builds or validates CVSS:4.0 vector strings, using the pinned Python cvss library (Red Hat Product Security).

  25. 2725
    P

    Pi extension showing subscription usage for supported model providers.

  26. 2726
    P

    OpenRouter multimodal tools for Pi — search, fetch, image gen, vision, video, PDF, TTS, STT

  27. 2727
    S

    Pi-native Compound Engineering package for iterative development workflows

  28. 2728
    P

    Lightweight CTF framework for Pi with an async Python engine, native pwn MCP, category playbooks, evidence, subagents and goal loop.

  29. 2729
    P

    Gemini LLM provider for Pi/GSD via A2A protocol with MCP tool bridge

  30. 2730
    P

    Interactive single/multi-question UI tool for PI coding agent.

  31. 2731
    P

    Web search and URL fetching for the tlh coding agent (Exa provider)

  32. 2732
    P

    Trigger broker and picker registration adapter for pi-interaction runtime

  33. 2733
    P

    pi bridge for the beyin second brain (JS CLI): session-start knock (ingest, dream) and one `beyin context` injection per session and cwd, plus a serial-tool-call nudge. Skills and tools live in the vault, not here.

  34. 2734
    P

    Review current pi work in a new branch with conversation context

  35. 2735
    P

    Config-driven typed classification for Pi via Microsoft Decision-1, OpenAI Decisions, Perplexity, TypeSafe Jev and Cloudflare Clef, with selective fallback and explicit comparisons.

  36. 2736
    P

    Automatic execution approval review for Pi, inspired by Codex's "Approve for me". Reviews tool execution with rules and Pi's current model, requesting user approval when needed.

  37. 2737
    U

    Terminal UI enhancements, styled bash tool rendering, and visual presentation components for Pi

  38. 2738
    P

    Pi extension that adds MiniMax, StepFun, and Command Code model providers plus a TypeSafe Jev decision tool

  39. 2739
    P

    A pi extension for loop

  40. 2740
    A

    A subagent tool supporting context compaction and nested subagents — user-customizable models (with round-robin) and tool policies, extensible via hooks.

  41. 2741
    P

    OpenAI Responses compaction, Fast mode, prompt promotion, and optional apply_patch for Pi

  42. 2742
    P

    Suggests your next prompt in grey after each reply; Tab takes it

  43. 2743
    P

    Agent sessions call each other across machines on your Tailscale network, in Pi and Claude Code.

  44. 2744
    R

    reddb.io internal plugin - maintainer-only skills for operating the red-skills repository.

  45. 2745
    P

    Pi provider for any compatible endpoint: log in with a base URL, API key, and protocol, and refresh its models automatically

  46. 2746
    P

    Bilingual developer companion & declarative Prompt Optimizer for Pi — Mode as Configuration, Dual-Core Golden Defaults, Zero-Residue First-Run Handshake, and S-Tier Prompt Architect.

  47. 2747
    P

    Generate AGENTS.md contributor guides for Pi repositories.

  48. 2748
    P

    Pi extension package for toggling OpenAI Fast Mode inference

  49. 2749
    P

    Make every model write the way you want, with consistent prose rules and styles you can switch instantly without restarting Pi.

  50. 2750
    P

    Keeps long Pi sessions cheap without losing anything: folds older output when the savings pay for the rewrite, prepares summaries in the background, and recalls originals byte for byte. Zero config.