Saket Aryan
47ce17c21b
Review found the contract documented but not implemented, and one client path missed entirely. AsyncMemoryClient's custom-client branch still carried the old literal header dict, so `AsyncMemoryClient(client=...)` sent no surface identity at all — the exact asymmetry this work set out to remove. Both custom-client branches also used a blanket headers.update(), which overwrites. That is the one code path where an outer layer's identity can physically be present, and it was the one path that erased it. They now check-then-set the identity headers and append to an existing client stack, which is what set-once and append-only were supposed to mean. AGENTS.md claimed a plugin calling the Python SDK produces `mem0-plugin/0.3.1, mem0-python/2.0.19`. Nothing in the repo sets the env vars that would make that happen, so the concatenation was unreachable. Replaced with the three ways an integration can actually declare itself, in preference order. memory_core's comment said the backend reads X-Mem0-Source. That is only true from the platform release shipping alongside this, and a reader would otherwise trust it and build header-only attribution that silently does nothing — which is how vercel-ai-sdk was written in the first cut. Corrected in all seven copies, and the body value is what makes attribution work against either backend. mem0-ts hardcoded SDK_VERSION = "3.1.8" while the repo already injects __MEM0_SDK_VERSION__ via tsup, the same mechanism telemetry.ts uses. The hardcode was correct only until the next release bump. Dropped both `as never` casts in pi-agent. They suppressed an excess-property error but also disabled checking of every other option at those call sites, so a typo in filters or threshold would have compiled. SearchMemoryOptions now declares `source` instead. Stack truncation cut mid-identifier, leaving a fragment that parses as a real client name. It now drops whole entries. Claude-Session: https://claude.ai/code/session_01C7tEmH86HAr7GoAAKCEHZb
@mem0/pi-agent-plugin
Persistent semantic memory for Pi Agent, powered by Mem0.
This extension gives Pi Agent long-term memory that persists across sessions, projects, and devices. Memories are automatically captured from conversations and can be searched and managed through slash commands and an agent-accessible tool.
Current package version: 0.3.0. Shared redaction and lifecycle utilities come from agent-plugin-core; Pi keeps its own tools and scopes.
Sidekick is available only in the Claude Code plugin.
Features
- Automatic memory capture: learns from every conversation (both user and assistant messages)
- Semantic search: find memories by meaning, not just keywords
- Scoped memory: project, session, or global scope
- Monorepo-aware: uses git root for project detection, consistent app_id across subdirectories
- Confirmation dialogs: destructive commands ask before acting
- 6 slash commands: essential memory management from the command line
- Agent tool:
mem0_memorytool lets the agent search and store memories autonomously
Setup
1. Get an API key
Sign up at app.mem0.ai and copy your API key.
2. Install
pi install npm:@mem0/pi-agent-plugin
3. Configure
Set the API key as an environment variable:
export MEM0_API_KEY="m0-your-key-here"
Or create a config file at ~/.pi/agent/mem0-config.json:
{
"apiKey": "m0-your-key-here",
"userId": "your-username",
"autoCapture": true,
"defaultScope": "project",
"searchThreshold": 0.3
}
Environment variables (MEM0_API_KEY, MEM0_USER_ID) override the config file.
searchThreshold (default 0.3) is the minimum similarity score (0–1) a memory must reach to count as a match for /mem0-search and /mem0-forget. It is passed to the mem0 search API (along with reranking for higher-precision ordering), so a query with no sufficiently similar memory reports no match instead of returning the closest unrelated memories. Raise it to be stricter; lower it if relevant results are missed.
Commands
| Command | Description |
|---|---|
/mem0-remember <text> |
Store a memory verbatim (no inference) |
/mem0-forget <query> |
Search and delete memories (with confirmation) |
/mem0-search <query> |
Semantic search across memories |
/mem0-tour [scope] |
Browse all memories grouped by category |
/mem0-scope <scope> |
Change default scope for this session |
/mem0-status |
Connection health, identity, and memory count |
Skills
The plugin includes 6 skills that guide the agent on how to use each capability:
| Skill | Purpose |
|---|---|
context-loader |
Pre-fetch relevant memories at session start |
remember |
Store facts with category classification |
search |
Quick semantic search with compact results |
forget |
Delete memories with confirmation |
tour |
Full memory walkthrough by category |
status |
Health check and diagnostics |
Memory Scopes
| Scope | Filters | Use case |
|---|---|---|
project |
user + app_id (git root) | Default. Project-specific knowledge |
session |
user + app_id + run_id | Recall restricted to memories saved with the current session ID |
global |
user only | All memories across all your projects |
Project scoping uses git rev-parse --show-toplevel to detect the repository root, so all subdirectories within a monorepo share the same memory pool.
Global tool operations require /mem0-scope global or defaultScope: "global" in plugin configuration. A model-supplied scope argument cannot enable cross-project access on its own. Empty or wildcard user, project, and session identities are rejected.
Automatic recall and capture
Before an agent response, the extension searches project memories and injects relevant results. After agent_end, automatic capture sends the user and assistant text supplied by Pi to Mem0 in project scope, regardless of the default scope selected for explicit commands. Captured text is redacted without the former 6,000-character per-message cutoff.
Automatic project writes do not attach run_id. Session-scoped recall applies to memories explicitly saved in session scope; it does not make project memories session-specific or automatically expire them. Recall queries and displayed tool results keep separate size limits.
Memory Categories
Memories are automatically classified into 10 general-purpose categories:
| Category | Description |
|---|---|
identity |
Personal details, background, self-descriptions |
preferences |
Likes, dislikes, habits, preferred approaches |
goals |
Objectives, aspirations, targets |
projects |
Ongoing work, initiatives, areas of focus |
decisions |
Choices made, rationale, trade-offs |
technical |
Technical knowledge, tools, configurations |
relationships |
People, teams, organizations |
routines |
Recurring patterns, workflows, schedules |
lessons |
Insights learned, mistakes to avoid |
work |
Professional context, role, responsibilities |
Architecture
pi-agent-plugin/
├── src/
│ ├── entry.ts # Extension entry point
│ ├── index.ts # Barrel exports
│ ├── commands.ts # 6 slash commands
│ ├── prompt.ts # System prompt injection (MEMORY_POLICY)
│ ├── types.ts # Shared interfaces and categories
│ ├── telemetry.ts # PostHog telemetry (batched, PII-safe)
│ ├── config/ # Config loading (~/.pi/agent/mem0-config.json)
│ ├── memory/ # Tool registration, scoping (git root), formatting
│ └── capture/ # Auto-capture from conversations (user + assistant)
├── skills/ # 6 SKILL.md files for Pi Agent
├── tests/ # Vitest unit tests
└── dist/ # Built output (ESM + DTS)
Development
pnpm install # Install dependencies
pnpm run typecheck # Type check
pnpm run test # Run tests
pnpm run build # Build (ESM + declarations)