Skip to main content

FrootAI mark

FAI Orchard Harvest MCP for Python

A Python, zero-runtime-dependency MCP surface for the governed Harvest pipeline.

PyPI · FAI Orchard · MCP guide

FAI Orchard Harvest product hero

Catalog or Harvest?

Need Package
Read published accelerator catalog records Orchard catalog source package frootai-mcp-orchard
Run repository discovery, extraction, retrieval, import, customization, or approval frootai-orchard-mcp — this package

Current version 0.1.1 is an early Harvest MCP surface. Confirm which handlers your installed release registers before relying on a specific pipeline operation.

Current registry truth: the tool registry declares nine handlers, including list_pending_reviews. The 0.1.1 CLI help banner still says eight; treat tools/list as authoritative until that banner is corrected in a future code release.

Five steps

1. Install or run in isolation

uvx frootai-orchard-mcp

Or install into a managed Python 3.10+ environment:

python -m pip install --upgrade "frootai-orchard-mcp[sdk]"

2. Connect an MCP client

{
  "servers": {
    "frootai-orchard-harvest": {
      "type": "stdio",
      "command": "uvx",
      "args": ["frootai-orchard-mcp"]
    }
  }
}

3. Discover, extract, and retrieve

discover_repo → extract_facts → retrieve_exemplars → list_plays

These free operations establish source identity and structured evidence before any import or customization.

4. Import and customize with entitlement

import_accelerator → customize_play → re_harvest

Import and drift operations require eligible entitlements. Policy overlays and three-way diffs must remain reviewable; no entitlement implies blanket approval to overwrite a target.

5. Separate generation from approval

list_pending_reviews is reviewer-gated. approve_play is founder-gated. Keep candidate generation, review, and final promotion as separate identities and operations.

See the governed Orchard lifecycle

Orchard builds, FAI Engine verifies, and Lab publishes evidence

Open the live FAI Orchard product lifecycle.

Nine-tool surface

Expand the Harvest pipeline tools and access boundaries
Tool Access Purpose
discover_repo Free Resolve and classify a source repository
extract_facts Free Extract structured architecture and IaC facts
retrieve_exemplars Free Find canonical reference Plays
list_plays Free Browse the slim Play catalog
import_accelerator Pro / import Run import and return signed artifact evidence
customize_play Pro / upgrade-to-play Apply organization policy and return a reviewable diff
re_harvest Pro / import Re-run at an updated source revision and report drift
list_pending_reviews Reviewer List candidates awaiting governance review
approve_play Founder Promote an authorized candidate

Embed with injected services

from frootai_orchard_mcp import TOOL_NAMES, create_server, run_stdio

server = create_server(
    services={
        # Inject repository, extraction, retrieval, artifact,
        # identity, telemetry, and governance adapters.
    }
)

print(TOOL_NAMES)
run_stdio(server)

The core is standard-library based and designed for hermetic testing. Missing service capabilities must return explicit errors rather than synthetic success.

Protocol resources and prompts

The Harvest surface exposes frootai://plays, frootai://policy-schema, frootai://auth/device, and frootai://exemplars/{id}. Guided prompts are import-rag, import-agent, customize-for-my-company, and audit-existing-infra. Prompts prefill instructions and never invoke tools on their own.

Transport options

# Stdio — default
uvx frootai-orchard-mcp

# Streamable HTTP + SSE
uvx frootai-orchard-mcp --http --port 8080

# Inspect without starting
uvx frootai-orchard-mcp --version
uvx frootai-orchard-mcp --help

The Python 0.1.1 CLI does not advertise the npm server's --ws option. Do not assume WebSocket parity; query the selected runtime's help and MCP capabilities.

Safety boundaries

  • Stdio is the smallest default process boundary.
  • FROOTAI_TOKEN and local identity state are resolved without returning raw credentials.
  • Treat source repositories and generated instructions as untrusted inputs.
  • Bind evidence to source revision, artifact identity, policy overlay, and approval receipt.
  • Validate generated infrastructure independently before deployment.
  • Keep reviewer/founder operations separate from automated import identities.

Verify and develop

Build the package and run repository-level Orchard/Harvest protocol and service-adapter tests. Every injected adapter should cover denial, timeout, malformed data, partial failure, retry, and rollback behavior.

License

MIT © 2026 FrootAI.

Metadata

Release files for frootai-orchard-mcp 0.1.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for frootai-orchard-mcp 0.1.1
File Size Uploaded Provenance Actions
frootai_orchard_mcp-0.1.1.tar.gz 66.3 kB Without provenance Details for frootai_orchard_mcp-0.1.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for frootai-orchard-mcp 0.1.1
File Interpreter ABI Platform Actions
frootai_orchard_mcp-0.1.1-py3-none-any.whl Python 3 none any Details for frootai_orchard_mcp-0.1.1-py3-none-any.whl

Total release size: 121.3 kB

Release files / frootai_orchard_mcp-0.1.1.tar.gz

Download URL frootai_orchard_mcp-0.1.1.tar.gz
Size 66.3 kB
Tags Source
SHA-256 checksum
How to use checksums
6a174b2b8d34d91dd2e6aa72d14be22f9fb62525071209e88f5eda7427b54572
BLAKE2b-256 checksum
How to use checksums
34cb71498e7ded5001fa96029c008515a07bc96c66c0b5751a0e509dd9d354e9
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.12.14

Release files / frootai_orchard_mcp-0.1.1-py3-none-any.whl

Download URL frootai_orchard_mcp-0.1.1-py3-none-any.whl
Size 55.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
97ad9d9ec3be486e98e68503a83375f18c7b722ac1b4f38441a61ca9ec4f5465
BLAKE2b-256 checksum
How to use checksums
f61bb7e58a27b702101a15e5ec73fcff50853c6c6d96831ad3f9a333e515b131
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.12.14

Security for frootai-orchard-mcp 0.1.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.1.1 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page