A local bridge server that lets Claude archive and retrieve agent configurations without network round trips. The MCP exposes publishing and retrieval operations for skills, prompts, tools, and guardrails, all backed by local storage with source tracing. Reach for this when you're managing multiple agent setups across different platforms and need a single source of truth that stays on your machine. The governance evidence layer gives you diffs and rollback capability, so you can see exactly what changed between exports. Built in Rust, runs over stdio, no telemetry or cloud dependencies. If you're already juggling agent configs manually or copy-pasting between surfaces, this gives you programmatic access to that workflow through Claude.
beam your agent config to any planet
What You Get • Install • How It Work • Ecosystem
AgentExport is a Rust workbench that exports your agent setup — skills, prompts, tools, modes — to every coding-agent surface in one shot. Configure once. Beam everywhere.
┌──────────────────────────────────────┐
│ LOCAL-FIRST ████████ 100% │
│ SOURCE-TRACEABLE ████████ 100% │
│ TYPING REQUIRED ░░░░░░░░ 0% │
│ VIBES ████████ ZERO │
│ FILLER│
└──────────────────────────────────────┘
One config, many agents. One beam, many platforms.
| Surface | What |
|---|---|
agent-export cockpit | Visual workbench. Pick targets, hit beam. |
agent-export studio | Live preview each agent's surface before you ship. |
agent-export policies | Per-agent guardrails. Same intent, different houses. |
agent-export skill catalog | One source of truth, fanned out to N hosts. |
agent-export cli | Reproducible exports. Diff what changed. Roll back clean. |
[!IMPORTANT] Local-first by default. No silent telemetry. No cloud round-trip. Your data stays on your machine until you explicitly ship it somewhere.
git clone https://github.com/xiaojiou176-open/agent-export.git
cd AgentExport
# follow the per-stack quickstart in INSTALL.md or docs/
Three commands. No curl | sh. No login. Read what you run.
Install break? Open your favorite agent and say "Read AGENTS.md and bootstrap AgentExport for me." Agent fix own brain. Long version: docs/.
The repo is seven layers — exactly the seven commits in git log. New work goes in as small named PRs. No 50-file mystery commits.
| Layer | What |
|---|---|
chore: scaffold | License, governance, hygiene gates, CI scaffolding. |
feat(core) | The primary engine. The reason AgentExport exists. |
feat(modules) | Packages, adapters, services, plugins. The second floor. |
feat(contracts) | Schemas, configs, public boundaries. Other code talks here. |
test: | Receipts. Everything in this layer must run. |
feat(ops) | Scripts, infra, CI helpers, build glue. |
docs: | Public docs surface. The pretty face. |
git log reads like a building floor plan. Look once, know the whole shape.
AgentExport lives in the *Me family: three personal tools. command-style names, stupid-simple jobs.
| Repo | What |
|---|---|
| AgentExport (you here) | beam your agent config to any planet |
| BrewMe | wake up, news already brewed |
| AgentGangGang | five AIs in one sidebar |
Cross-family taste: AgentExport · BrewMe · OpenVibeCoding · proofyard.
If AgentExport saves you a click, an hour, or a headache — star costs zero. Fair trade. ⭐
MIT — small print, big freedom.
io.github.ericm1018/skillfm-llm-cost-optimizer-openai-anthropic-usage
io.github.mikerawsonnz/llm-orchestration-agent
io.github.mikerawsonnz/authenticated-llm-agent
labforgedev/copilot-memory-mcp
csoai-org/agent-prompt-injection-firewall-mcp
io.github.mikerawsonnz/authenticated-multi-llm-agent