Get This Tool
Persnally
Summary
You paste your stack into Claude, then paste it again into Cursor, then again when you open a new ChatGPT thread — because every tool starts blank and the vendors have no incentive to change that. Persnally is a local daemon that learns your context once from your AI history and serves it to every MCP-compatible tool the moment a session starts.
The setup is a single npm install and an import command that reads your Claude exports, Claude Code sessions, and git repos, then builds a structured, evidence-linked profile in ~/.persnally. From there, a local daemon serves that profile over MCP so Claude, Cursor, Copilot, and other connected tools load your conventions and preferences at session start without you repeating them. Per-client scopes let you decide exactly what each tool can see. The architecture is the trust model — context never leaves your machine, deletions erase source events and rebuild all derived views with no residue. Where it breaks: any tool that does not speak MCP stays out of the loop, and teams using proprietary AI integrations without MCP support get nothing from this.
Bottom line: Pick this if you switch between three or more MCP-compatible AI tools daily and are tired of re-explaining your stack — skip it if your primary tools lack MCP support, because the daemon has nothing to connect to.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Community Benchmarks Community
Sign in to submit a benchmarkNo community benchmarks yet. Be the first to share a real-world data point.
Pros
Sign in to edit- Learns your stack, conventions, and preferences from existing AI history on import, so you stop re-explaining the same context at the start of every new session across different tools.
- Serves context over MCP — the protocol already adopted by Claude, Cursor, Copilot, and others — which means no custom plugin per tool; one daemon serves all connected clients.
- Per-client scopes let you control exactly what each AI tool is allowed to read, so a scoped agent gets a narrower view than your primary coding assistant without manual prompt filtering.
- Provenance-complete profile means every inferred trait links to the source events that produced it, so you can audit what the tool thinks it knows and why rather than treating it as a black box.
- Deletion erases source events and rebuilds all derived views from zero — no soft deletes or hidden residue — which means removing a topic actually removes it from everything downstream.
Cons
Sign in to edit- Any AI tool that does not implement MCP cannot call get_context and receives no context at all — teams whose primary tools lack MCP support get zero benefit from the daemon, and at that point they are back to manual context pasting or switch to a tool with a native memory feature built into the vendor's platform.
- The profile is only as accurate as the history you can import; the page names Claude exports, Claude Code sessions, and git repos as sources, so workflows living in other tools — Slack threads, Jira tickets, proprietary IDEs without export — contribute nothing until an import path exists for them.
- Running a persistent local daemon adds a process to maintain across machine restarts and OS updates; teams on locked-down corporate machines or shared development environments without npm global install rights hit a setup wall before the tool does anything useful.
Community Reviews
Sign in to write a reviewNo reviews yet. Be the first to share your experience.
About
- Platforms
- macOS, Linux, Windows (Node 20+)
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- 2026-06-30T12:18:19.504Z
Best For
Who it's for
- Developers who switch between several AI assistants
- Users who want full local control over personal data
- Anyone building or using MCP-compatible AI clients
What it does well
- Maintain consistent context across multiple AI coding and chat tools
- Avoid repeating project details and preferences in every session
- Audit or delete specific topics from the learned profile
Integrations
Discussion Community
Sign in to commentNo discussion yet. Sign in to start the conversation.
Compare Persnally
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeCommunity Notes & Tips Community
Sign in to contributeBe the first to contribute. General notes, observations, gotchas, and tips from people who use this tool day-to-day.
Frequently Asked Questions
- Is Persnally free?
- Yes — Persnally is fully free to use. There is no paid tier.
- Is Persnally open source?
- Yes. Persnally is open source.
- Can I self-host Persnally?
- Yes. Persnally supports self-hosting on your own infrastructure.
- What platforms does Persnally support?
- Persnally is available on: macOS, Linux, Windows (Node 20+).
Hours Saved & ROI Stories Community
Sign in to contributeBe the first to contribute. Concrete time/cost savings, with context. e.g. "Cut my code review backlog from 4h to 45m per week."
Curated lists that include this category
Every AI coding and chat tool you open meets you as a stranger. Persnally solves that by running a local daemon that ingests your AI conversation history and git activity, derives a structured profile of your preferences, stack, and decisions, and serves that profile on demand over MCP. The workflow is three steps: import your history with one command, let the local process build the model, then every MCP-connected tool calls get_context at session start and receives your context without you typing a word.
The differentiating design choice is provenance. Every claim the profile makes links back to the specific events that produced it — you can ask ‘why does it think this?’ and get a real answer pointing to exact imports, decisions, or conversations. Deleting a topic erases the source events and rebuilds all derived views from scratch; the docs describe this as leaving no tombstones, no residue. That is an architectural guarantee, not a privacy policy.
Persnally fits developers who rotate across multiple AI assistants and want context continuity without manual prompt engineering each session. It also fits anyone who will not hand personal AI history to a cloud service — the vendor states the profile lives in ~/.persnally and is never sent to their infrastructure. The ceiling is MCP coverage: tools that do not implement MCP cannot call get_context, so any part of your workflow outside the MCP ecosystem stays disconnected. Per-client scopes let you restrict what individual tools — including scoped agents — are allowed to read, which matters when you connect tools with different trust levels.
Installation is via npm (npm i -g persnally), the daemon stores data as SQLite and JSON files under ~/.persnally, and the license is FSL-1.1-MIT. The vendor states no paid tiers and no cloud offering exist. Supported import sources named on the page include Claude conversation exports, Claude Code sessions, and git repos; supported MCP clients named include Claude, ChatGPT, Cursor, Copilot, Zed, Windsurf, and Cline.