Coact Publish

coact publish — Python capability → Claude integration

coact publish ships a capability to a chatbot host. Today it has one target, claude-local-mcpb: a Claude Desktop .mcpb Desktop Extension that runs a local stdio MCP server. The MCP server itself is built by py2mcp; coact writes the packaging.

When to use

• The user has Python functions (or a skill carrying a coact: mcp: block) and wants them usable inside Claude as tools.
• They ask to "make / build / package / deploy" a Claude connector, plugin, extension, .mcpb, or MCP server from local code.

How

Always preview first (writes nothing):

coact publish my.package.module:my_func another.module:other_func --dry-run

Then build the bundle:

coact publish my.package.module:my_func --name my-tools --dest ~/Downloads
# → ~/Downloads/my-tools.mcpb

Sources accepted (mix freely): module:function refs, a skill directory / SKILL.md (its coact: mcp: block supplies the refs), or — from Python — live callables and a prebuilt IntegrationSpec.

From a natural-language description (opt-in LLM)

To go from an English description to a draft integration (proposed tools with inferred input schemas), use coact describe — the only LLM-using path here (the module:function → .mcpb path stays LLM-free). Generation routes through aix (multi-provider) by default; --llm picks a model.

coact describe "a connector that looks up the weather for a city and converts currencies"
# → renders a draft IntegrationSpec: proposed tools, each marked "proposed (no handler)"

from coact import integration_spec_from_description, publish

spec = integration_spec_from_description(
    "expose os.path.basename and os.path.dirname as tools", name="paths"
)
# tools the description bound to existing code become runnable refs:
publish(spec, name="paths", dest="dist")   # works iff spec.runnable_refs()

The draft is a design artifact: tools are proposed (no importable handler) unless the description named existing code. Bind each proposed tool to a real module:function handler (write the code, or point at existing functions) before publish can build a runnable .mcpb — coact writes the design, you own the code. Needs coact[nl] (oa, aix), imported lazily only on this path.

Python API:

from coact import publish, integration_spec_from

publish(["mypkg.tools:summarize", "mypkg.tools:translate"],
        name="text-tools", dest="dist", author="Me")

Install the result: Claude Desktop → Settings → Extensions → Install Extension… (or double-click the .mcpb). The extension runs on the user's machine and needs a Python with py2mcp + fastmcp importable.

Remote claude.ai connector (Streamable-HTTP + OAuth 2.1)

A claude.ai custom connector is a remote MCP server reached from Anthropic's cloud over HTTPS + OAuth — a different surface from the local .mcpb. Scaffold one (a hosted service you deploy) with the claude-remote-connector target:

coact publish mypkg.tools:summarize --target claude-remote-connector \
  --name my-conn --dest ./out \
  --connector-url https://my-conn.example.com --idp-issuer https://my-idp.example.com
# → ./out/my-conn-connector/  (server/app.py, connector_config.json, requirements.txt,
#                              DEPLOY.md, Dockerfile)

from coact import publish_remote
publish_remote(["mypkg.tools:summarize"], name="my-conn", dest="out",
               connector_url="https://my-conn.example.com",
               idp_issuer="https://my-idp.example.com",
               required_scopes=["mcp:read"])

It scaffolds an OAuth 2.1 resource server (validates a managed IdP's JWTs via py2mcp.http.mk_http_app; never issues tokens; audience-bound per RFC 8707). Omit --connector-url/--idp-issuer to scaffold with fill-in placeholders + a loud warning. Then follow the generated DEPLOY.md: set your IdP, run behind TLS (uvicorn server.app:app), and add the HTTPS URL as a custom connector in claude.ai. Needs py2mcp>=0.1.4 + fastmcp + uvicorn where the service runs.

Key distinctions (don't conflate)

• Local .mcpb (claude-local-mcpb): stdio, no OAuth, runs on the user's machine.
• Remote connector (claude-remote-connector): a hosted MCP server reached from Anthropic's cloud over HTTPS + OAuth — public, multi-user, you deploy it.
• A connector is connectivity (tools). A Skill (SKILL.md) is procedural knowledge. They are complementary.

Limitations (current)

• Targets: claude-local-mcpb and claude-remote-connector. Claude Code plugins, ChatGPT Apps, and Gemini are planned (the registry is open-closed).
• Tools are referenced by module:function; the functions must be importable where the server runs (dependency vendoring is a future refinement).
• Background: misc/docs/CHATBOT_INTEGRATION_LANDSCAPE.md.