Pi supports context files, project settings, project-local extensions, Agent
Skills, and package-managed extensions. Pi core intentionally does not include
built-in MCP; the starter package uses pi-mcp-adapter.
Official references:
The repository includes a Pi starter package at .pi/. Copy that folder to
your workspace root, then render the copied package. The .pi/render-starter
script is a convenience: with a single --repo list such as
--repo my-app shared-lib, it infers the workspace root from the copied .pi/
folder, fills path and repository placeholders, renders the project extension
constant, and validates that each requested repository exists before you start
or reload Pi from the workspace once. If you prefer not to run the renderer,
make those same replacements by hand and verify that no placeholder tokens
remain.
The package contains:
.pi/settings.json - project settings that load the local extension, local
skills, and npm:pi-mcp-adapter..pi/extensions/agents-remember-start.ts - project-local
before_agent_start extension that injects the mandatory first-action
directive..pi/mcp/agents-remember-settings.json - Agents Remember MCP authority
settings..pi/mcp.json - Pi project override consumed by pi-mcp-adapter..pi/skills/ - Agents Remember skills in Pi’s project skill root.After the restart or reload, invoke:
c-13-install-and-onboard
That skill runs or verifies runtime_install() and then handles memory,
onboarding, and providers.
The starter package uses a project-local .pi/extensions/agents-remember-start.ts
extension. Pi runs before_agent_start before an agent turn and allows the
extension to inject the mandatory first-action directive into the system prompt.
Do not overwrite a repository’s root AGENTS.md just to wire Agents Remember;
root AGENTS.md is project-specific. Pi also supports global context under
~/.pi/agent/AGENTS.md, but the starter package does not need it for first-run
setup.
If you edit the extension while Pi is already running, use /reload.
After the restart or reload, the copied c-13-install-and-onboard skill runs or
verifies:
runtime_install()
Pi loads skills from project .pi/skills, project .agents/skills, global
~/.pi/agent/skills, global ~/.agents/skills, settings entries, package
entries, and repeated --skill <path> flags. The starter package keeps skills
in .pi/skills and also lists skills in .pi/settings.json to make the
project package self-contained.
Pi can discover SKILL.md directories recursively, but skill names should still match their parent folder for clean validation.
Do not run skills_install() for first-run setup. It remains available for
manual maintenance and non-package installs.
Pi core has no built-in MCP client. The starter package loads
npm:pi-mcp-adapter from .pi/settings.json; Pi installs missing project
packages automatically on startup.
pi-mcp-adapter can read shared project .mcp.json, but this starter keeps the
Pi package self-contained with the documented Pi-owned project override
.pi/mcp.json:
{
"mcpServers": {
"agents-remember": {
"command": "uvx",
"args": [
"--refresh-package",
"agents-remember-mcp",
"agents-remember-mcp@latest",
"--config",
"<PATH/TO/YOUR/PROJECTS_FOLDER>/.pi/mcp/agents-remember-settings.json"
]
}
}
}
Use /mcp to inspect adapter status or /mcp setup when you want Pi to adopt
existing host-specific MCP configs.