A Python execution engine for AI agents. It runs browser and API work as
explicit steps, records what every step did, and replays from the step that
failed — instead of re-running the whole job.
The current public inventory is 482 registry-backed modules across 89
catalog categories, including triggers, queue modules, workflow versioning,
metering hooks, browser automation, API calls, data transforms, verification,
files, and crypto.
Every run produces an execution trace and state snapshots. If step 3 fails, replay from step 3 — no re-running the whole thing.
Usage
# Run a built-in recipe
flyto recipe site-audit --url https://example.com
# Run your own YAML workflow
flyto run my-workflow.yaml
# List all recipes
flyto recipes
Flyto2 Core exposes the same deterministic runtime through several supported interfaces rather than separate execution engines:
Python — import the module registry and workflow engine directly.
YAML workflows — compose registered modules into replayable procedures.
Execution API — use the authenticated /v1/* HTTP routes for local integrations.
MCP — expose the registry through stdio or Streamable HTTP to AI clients.
Module contract — every executable action is defined by registry metadata and a bounded parameter schema; host-only primitives such as capability.invoke additionally require opaque runtime authority and cannot be activated by serialized workflow data alone.
Flyto2 Core runs with safe defaults. Optional providers, browsers, verification services, and connectors are enabled explicitly through package extras and documented environment variables; secrets stay in runtime environment/credential stores rather than workflow YAML. Start from .env.example for supported settings and see Operations for runtime/deployment guidance.
Host-injected runtime capabilities are deliberately different from configuration: a workflow cannot enable capability.invoke by setting an environment variable or serialized parameter. The trusted host must inject the execution-scoped opaque authority for that run.
Execution Trace — structured record of every step: input, output, timing, status
Replay — re-execute from any step with the original (or modified) context
Breakpoints — pause execution at any step, inspect state, resume
Evidence Snapshots — full state before and after each step boundary
Data Lineage — track data flow across steps, build dependency graphs
Timeout Guard — configurable workflow-level and per-step timeout protection
Architecture
CLI, MCP, HTTP, Python, and packaged recipes converge on the same workflow
engine, module registry, policy, trace, evidence, and replay boundaries. Start
with the Technical Whitepaper, then use the
Architecture Map and exhaustive
source reference for implementation detail.
The shared product contract, flyto.product-contract.v1,
defines the Flyto2 promise: Turn AI work into verified, replayable procedures.
Package
Responsibility
flyto-ai
Understand, route, and govern new work and provider use.
flyto-blueprint
Store, learn from, and score reusable procedures; it never executes them.
flyto-core
Validate schemas, execute and replay deterministically, and emit evidence.
flyto-core is a standalone execution package; it does not require the other
packages to execute a workflow or produce evidence.
The canonical PyPI and MCP registry description is: The open-source execution engine for AI agents. 482 modules, MCP-native, triggers, queue, versioning, metering.
Contributing
We welcome contributions! See CONTRIBUTING.md for guidelines.
Report security vulnerabilities via [email protected].
See SECURITY.md
for the security policy and the environment variables that define the filesystem
and outbound-network boundaries.
SECURITY_STATUS.md
lists every published advisory with its severity, affected range, fixed-in
version, and the regression test that covers it. Two boundaries are enforced
registry-wide by tests that fail the build — every module taking a
caller-supplied path must reach the filesystem sandbox helper, and every module
taking a caller-supplied URL or host must reach an SSRF guard — so coverage is a
CI property rather than a convention.
Also known as: open source AI agent framework for production workflows · Python AI workflow automation with Playwright · MCP server automation with trace and replay · browser automation that can resume from a failed step
Noch keine Tool-Liste. Remote-Server erhalten eine nach dem ersten Health-Check.
FAQ
Fragen
Wie installiere ich Flyto Core?
Kopieren Sie auf dieser Seite die stdio-Konfiguration für Cursor, Claude Desktop, Claude Code oder VS Code. Sie nutzt die Installation des Herausgebers, keinen PlugMCP-Proxy.
Benötigt Flyto Core einen API-Schlüssel?
In diesem Eintrag ist keine erforderliche Umgebungsvariable aufgeführt.
Welchen Transport verwendet Flyto Core?
Flyto Core läuft lokal über stdio.
Wann wurde Flyto Core zuletzt geprüft?
Zuletzt geprüft: 2026-11-12. AI said it finished. Flyto2 shows the proof.