From attune-author
Generate full documentation from source code with the 3-stage AI pipeline (outline, write, review). Triggers on: write docs, generate API reference, create README, AI doc generation.
How this skill is triggered — by the user, by Claude, or both
Slash command
/attune-author:author-docsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Generate full documentation from a source file using the
Generate full documentation from a source file using the 3-stage AI pipeline: outline → write → review. Output is production-ready API references, guides, or READMEs.
ANTHROPIC_API_KEY must be set in the environment[ai] extra must be installed:
pip install 'attune-author[ai]'Use AskUserQuestion to gather inputs:
question: "What kind of documentation do you need?"
header: "Doc Type"
options:
- label: "API reference"
description: "Function/class signatures, args, returns, examples"
- label: "User guide"
description: "How-to walkthrough for end users"
- label: "README"
description: "Project overview, install, quick start"
- label: "Module overview"
description: "Architecture and component descriptions"
Then ask for audience (developers, beginners, ops) and
the target source path.
author_docs with target,
doc_type, audience, and optional output_path.output_path is set, confirm the file was written.
Otherwise show the content inline.For larger jobs (multi-file modules, full directories),
delegate to the doc-writer subagent instead of calling
the tool directly. The agent has more turns to work with
and can iterate on multiple files.
Stages: outline → write → review
Output: docs/api/auth.md (4827 chars)
npx claudepluginhub smart-ai-memory/attune-docs --plugin attune-authorProvides CDSS development patterns for drug interaction checking, dose validation, clinical scoring (NEWS2, qSOFA), and alert classification integrated into EMR workflows.