avtc-pi

An extension suite — deterministic feature development, subagent delegation, working-memory, behavioral learning, parallel-work guardrails, durable decisions, notifications, and more.

Packages

Package details

extensionskill

Install avtc-pi from npm and Pi will load the resources declared by the package manifest.

$ pi install npm:avtc-pi
Package
avtc-pi
Version
1.1.1
Published
Jul 17, 2026
Downloads
1,072/mo · 1,072/wk
Author
avtc
License
MIT
Types
extension, skill
Size
7.2 KB
Dependencies
14 dependencies · 0 peers
Pi manifest JSON
{
  "extensions": [
    "../avtc-pi-featyard/index.ts",
    "../avtc-pi-subagent/index.ts",
    "../avtc-pi-todo/index.ts",
    "../avtc-pi-portrait/index.ts",
    "../avtc-pi-parallel-work-guardrail/index.ts",
    "../avtc-pi-user-decisions/index.ts",
    "../avtc-pi-ui-components/index.ts",
    "../avtc-pi-subagent-ui-bridge/index.ts",
    "../avtc-pi-unstuck/index.ts",
    "../avtc-pi-zai-continue/index.ts",
    "../avtc-pi-ask-user-question/index.ts",
    "../avtc-pi-notification/index.ts"
  ],
  "skills": [
    "../avtc-pi-featyard/skills"
  ]
}

Security note

Pi packages can execute code and influence agent behavior. Review the source before installing third-party packages.

README

avtc-pi

An extension suite — deterministic feature development, subagent delegation, working-memory, behavioral learning, parallel-work guardrails, durable decisions, notifications, and more.

This is a bundle package with no logic of its own. Installing it brings the entire suite in one step: its pi.extensions lists every avtc-pi extension, and each loads (guarded against double-registration via its own idempotency flag) so the bundle coexists safely with any extensions you also installed standalone.

What's in the suite

Extension What it does
avtc-pi-featyard Predictable, deterministic feature development — deep upfront design, configurable review and verification rigor, and auto-agents draining a backlog
avtc-pi-subagent A subagent tool supporting context compaction and nested subagents — user-customizable models (with round-robin) and tool policies
avtc-pi-todo A working-memory plan the agent manages — breaks multi-stage work into items when needed
avtc-pi-portrait Builds a behavioral portrait from your session history — learns your corrections and injects them into the system prompt
avtc-pi-parallel-work-guardrail Lets you block or approve agent-called git operations that disrupt parallel work
avtc-pi-user-decisions Captures user decisions and re-injects them into the system prompt after compaction and into subagents
avtc-pi-settings-ui A schema-driven settings UI — define your settings, get a tabbed modal, a /<name>:settings command, and layered persistence
avtc-pi-ui-components A dialog coordinator preventing dialogs from rendering over each other, plus a select-with-note component
avtc-pi-subagent-ui-bridge Lets extensions' dialogs from nested subagents render in the root session
avtc-pi-unstuck Auto-continue on empty model responses + configurable timeouts for bash and search tools
avtc-pi-zai-continue Auto-continue after Z.ai usage-limit resets, with a live countdown and retry suppression
avtc-pi-ask-user-question A question tool for the agent — single/multi-select or free-text, with subagent forwarding and attention alerts
avtc-pi-notification Bell and Telegram notifications on agent completion, errors, or attention needed — only fires when you're away
avtc-pi-logger File-based logging library for pi extensions — date-partitioned with rotation

Installation

pi install npm:avtc-pi

This installs the whole suite. Each extension can also be installed individually (pi install npm:avtc-pi-featyard, etc.) if you only want part of it.

Configuration

Each extension has its own settings, edited via its own /<name>:settings command (e.g. /fy:settings, /todo:settings). See each extension's repo for its configuration reference.

Developed with Z.ai — get 10% off your subscription via this referral link.

License

MIT