Skip to content

docs: add Python SDK quickstart tutorial#12

Open
ekline[bot] wants to merge 1 commit into
mainfrom
docs/tutorial-gap-review-2026-03-25
Open

docs: add Python SDK quickstart tutorial#12
ekline[bot] wants to merge 1 commit into
mainfrom
docs/tutorial-gap-review-2026-03-25

Conversation

@ekline

@ekline ekline Bot commented Mar 25, 2026

Copy link
Copy Markdown
Contributor

Context

This tutorial addresses the #1 gap identified through a tutorial coverage audit that included competitor analysis, persona needs assessment, and feature coverage review.

The problem: Fastn has a full-featured Python SDK (fastn-ai on PyPI) with 250+ connectors, semantic tool discovery, and LLM agent integration — but zero tutorial coverage for developers who want to use it programmatically. All existing tutorials are UI/flow-builder focused. Every major competitor (Composio, Nango, Paragon, Merge) has an "install → authenticate → first API call in 5 minutes" tutorial as a core onboarding path.

Top 3 gaps identified

Rank Priority Gap Source Personas served
1 High Python SDK quickstart — no tutorial for install, auth, first tool call All 4 competitors have this; developer persona blocked AI engineers, backend developers
2 High MCP server setup with Claude Desktop/Cursor — no developer-focused tutorial 3/4 competitors cover this; AI engineer persona blocked AI engineers
3 Medium CLI getting started — no tutorial for fastn login, fastn connector, fastn flow Developer persona needs; new feature without tutorial Backend developers

What this PR adds

  • New file: getting-started/sdk-quickstart.md — A 5-minute tutorial that walks developers through:
    1. Installing the fastn-ai package
    2. Configuring credentials (env vars, constructor params, config file)
    3. Sending a Slack message via fastn.slack.send_message()
    4. Error handling with SDK exceptions
    5. Semantic tool discovery with get_tools_for()
    6. Building an LLM agent with OpenAI integration
    • Includes troubleshooting table for common issues
  • Updated: SUMMARY.md — Added navigation entry under "Getting started"

Research sources

  • Competitors audited: Composio (docs.composio.dev), Nango (docs.nango.dev), Paragon (docs.useparagon.com), Merge (docs.merge.dev)
  • Personas served: AI engineers integrating Fastn with LLM agents, backend developers building programmatic integrations
  • Code verified against: fastn-sdk/python/fastn/client.py, config.py, connector.py, exceptions.py, _constants.py, pyproject.toml, and example files

Remaining gaps (follow-up work)

Quality review

  • Completeness: All required tutorial sections present (intro, prerequisites, 6 steps, summary, next steps, troubleshooting)
  • Technical accuracy: All code examples verified against SDK source — method signatures, env var names, package name, exception classes all confirmed
  • Style compliance: Active voice, second person, present tense, verb-first procedures, no banned phrases, sentence case headings
  • Links: All 3 internal links verified to resolve to existing files
  • Grade: A- (only note: H3 headings under H1 match existing repo convention)

How to verify

  1. Check out this branch
  2. Run the GitBook dev server or preview
  3. Navigate to Getting started → Python SDK Quickstart
  4. Walk through the tutorial steps with a Fastn account and verify the code examples work

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants