01Why it exists
Writing, editing, review and publishing are often handled by different people. Chat-only handoffs easily lose versions, authority and measurement context.
AI Content Operations
Help the next person pick up where you left off.
A five-module workflow connecting discovery, production, platform adaptation, release control and measurement feedback.
970 / Open methodA method to take with you
02Current state
The public kernel defines five modules and six artifact types while local adapters retain brand, platform and account boundaries. GitHub main commit 3a78a8cb1416c68c7fe545148bed90ecb273ea2f also adds a general delivery evidence check. It separates a done claim into six questions covering project identity, exact-revision checks, the delivery environment, recipient access, an important workflow with a relevant failure case, and the existing handoff record. It produces a result for human review; it does not deploy, publish, or prove live synchronisation, automatic triggers, live metrics or efficiency gains.
03How AI contributes
Content · Automation · Testing · Data analysis
04Public boundary
The public package contains no credentials, private paths, internal names, client or family data. Login, upload, scheduling and publication remain behind explicit human approval.
05Next
Continue testing whether the six delivery questions reduce revision mismatches and inaccurate handoff records in local content work, without treating N/A, unavailable and zero as the same state or claiming an efficiency gain in advance.
Sources
Update
From my AI voice to Chinese and English videos
The full edits passed my listening review after changes to the voice, pauses and transitions.
UpdateAI Content Operations adds a six-question delivery evidence check
A new offline Skill reviews project identity, revision checks, delivery, recipient access, experience and handoff records before done becomes a delivery claim.
UpdateAI Content Operations expands into a five-module public workflow
The public package now covers five modules, six artifact types, a fictional full lifecycle and offline validation.
Writing
Can an AI content workflow work on a new computer and account?
First, I completed the 970LAB AI content workflow on a Mac mini. Later, I automated publishing for Guoma970 on a MacBook Pro. Two separate projects, and the next steps toward reusable instructions.
WritingClaude stopped. My work kept moving.
Keep work records outside the AI account: where you got to, why you made a decision, and what comes next. Another AI can continue the work, but you still check the results and make the decisions.
WritingSomeone asked for a Skill, so I started packaging my workflow.
Packaging voice, video, and content-operations workflows, testing them on another Mac, and clarifying the inputs. A process record, with no download in this update.
WritingMy video editor preferred my AI voice to my own recording
I made Chinese and English videos with an authorised AI version of my voice, then asked the editor who knows my recordings to listen to the finished clips.
WritingAI finally sounded like me. Then I put it in a video.
A familiar voice was only the start. The pauses, delivery and full edit still needed a human listen.
WritingWhen AI Says It’s Done, Ask These Six Delivery Questions
Check the project, exact revision, delivery environment, recipient access, important workflow and handoff record before treating done as delivered.
WritingHow AI can carry a content team’s handoffs across writing, editing and publishing
The goal is not to replace every role, but to leave a usable artifact at every handoff.