
OSource-Manager
Updated 1mo agoby jgharbieh
See it · try it · keep it or kill it — with the verdict remembered.
Every repo, global CLI, MCP server and skill on your machine, in one place, with a journal.
The problem
You clone things. A repo to try, a CLI you installed globally once, an MCP server, a skill dropped into a dotfiles folder. Six months later there are sixty of them and no answer to the only questions that matter: what is this, do I still use it, and can I get rid of it?
Package managers track versions. They do not track your verdict. OSource-Manager does — it joins what's on your disk ↔ what upstream did ↔ do I still want this ↔ act on it, and it remembers the decision in an append-only journal so a six-month-old row can explain itself.
The funnel
Every tool moves through five states. Your verdict is owned by you and is never overwritten by a scan.
| 01 · Wanted | Tracked. May not even be on disk yet. | wanted |
| 02 · Trying | Giving it a shot — often in a Docker sandbox. | trying |
| 03 · Kept | Earned its place. | kept |
| 04 · Serving | Registered as an MCP server in ≥1 agent. | derived |
| 05 · Retired | Killed, with the reason kept forever. | retired |
Quickstart
pnpm install
pnpm build
node dist/cli.js setup # first-run: import everything found, offer self-registration
node dist/cli.js serve # web UI at http://localhost:7807
First run scans your configured dirs, package managers (npm -g, winget), agent configs and docker ps, and imports what it finds. Nothing leaves your machine.
What it does
| Trustworthy inventory | One row per tool, many installations. SSH / HTTPS / symlink variants collapse to one identity. A repo that ships a CLI and a skill is one row, not three. |
| Upstream intelligence | Release changelog since your version (GitHub) or latest published version (npm), ETag-cached, on-demand. No daemon. |
| README, rendered | GitHub renders hosted READMEs; local SKILL.md / AGENTS.md render through a built-in markdown pass — frontmatter, tables, code, images and video embeds. |
| Guarded updates | Fast-forward only. Dirty worktree, detached HEAD, or a diverged branch is refused before anything is touched — the checkout stays byte-identical. |
| Docker-sandboxed trials | Clone and run untrusted code without it ever touching your disk. See below. |
| MCP registrar | Add a server to Claude Code / Codex through their official CLIs, with a dry-run diff first and a real inverse. Registration is reversible. |
| The journal | Every mutation writes an event in the same transaction as the change — tracked, tried, updated vX→vY, registered, retired: reason. Your own notes interleave. |
Trying untrusted code, safely
Open-source software can carry malicious executables. The last thing you want is to install one and let it run arbitrary code on your machine. So a trial never touches your disk — it clones into a Docker volume, and the container that holds it is locked down.
Inspect mode (default) — read code nobody has audited, with nothing it can do:
Related servers

n8n
Updated todayby n8n-io
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.

mcp-server-fetch
OfficialUpdated 14d agoA Model Context Protocol server providing tools to fetch and convert web content for usage by LLMs

@modelcontextprotocol/server-everything
OfficialUpdated 14d agoMCP server that exercises all the features of the MCP protocol