CryptPad can provide encrypted collaborative editors inside another application, but its API is easy to mistake for a cloud document REST service. This skill helps your agent pick the supported interface and identify who must store documents, authorize users and manage keys.

Your agent gets practical guidance for embedding editors, working with existing CryptDrive documents, and diagnosing an instance. A small CLI checks public resources without executing server JavaScript, while a browser adapter handles save acknowledgements and concurrent key updates.

What You Get

Directory	Contents
`scripts/`	Read-only instance probe, offline link inspector, and contract tests
`references/`	Architecture, API contracts, operations, and dated primary sources
`templates/`	Integration plan, browser configuration, and callback adapter
`evals/`	Output-quality scenarios and separate trigger probes

Quick Start

From this directory:

python3 scripts/cryptpad.py probe --origin https://cryptpad.example.org
python3 scripts/cryptpad.py inspect-link < /secure/path/document-link.txt

Replace the example origin with your instance. A successful probe reports "ok": true and recognized config/loader endpoints; browser editing still needs separate verification. The offline inspector returns the application and whether a fragment exists, without exposing keys.

Triggers

“Can CryptPad edit documents stored by our application?”
“Build an embedded editor with reliable saves and key rotation.”
“Why does our CryptPad instance load but collaboration fail?”
“Export a CryptDrive document through my authorized browser session.”

Requirements

Python 3.10+; no Python packages or API key for discovery. Editing needs a browser, an instance permitting the required embedding, and authorized document access. Node.js runs the optional adapter contract test. This is not a headless CryptDrive CRUD client.