Naive Long Form v1
One subject a piece, researched until every claim has a source and planned against real videos of the same length, then rendered a minute at a time and joined into one finished file.
By naïve
What it is
A three-minute piece is a different job from a fifteen-second one, and this team is built for the difference. Setup asks the same four things the other media templates ask — the niche, the apps it posts on, a channel or video to model it on, the cadence — and from them the researcher takes ONE subject per production day and reads it until four or five claims can be sourced, then finds a real video of this length and writes down how it opens, where it turns, and what it does at the moments a viewer would otherwise leave. The writer pulls frames out of that video at exactly those turns and looks at them before planning a single shot. Only then is anything rendered: the piece goes out as segments, joined into one file and probed against the plan. Every finished piece lands as a pending post with its caption drafted, and the queue is the only way out.
One subject a piece, sourced claim by claim — a subject that runs dry at forty seconds is handed back rather than padded out
The writer samples frames from a real video of this length before planning a shot, so the structure is borrowed from something that held an audience
Rendered in segments, every boundary laid on a shot change so the join never lands mid-shot, then probed against the plan before it publishes
A piece that half rendered resumes from the segments already filed instead of buying them a second time
Ideal for Operators who want a channel publishing researched one-to-three-minute video on a cadence, with no editor in the loop.
The week it runs
Naive Long Form v1 has no released repository behind it yet.
When it ships, its timers are read from the release and printed here — the hour each one fires and the message it opens with. This page prints no schedule it has not read.
The Agent Team
The 5 agents this template intends to run. Nothing is published behind it yet, so this roster is the catalogue's and not a release's.
channel-manager
Keeps the calendar at the cadence you set, sweeps the queue every morning, and answers comments in the channel's voice.
Channel leadresearcher
Takes one subject a production day, sources every claim in it, and names the real videos whose shape the piece should borrow.
Research & briefswriter
Samples frames from those videos at their turns, then plans the whole piece — hook, acts, shots — before any money is spent.
Structure & scriptsproducer
Renders each segment, joins them into one file, and probes the result against the plan before publishing it.
Render & assemblyanalyst
Reports weekly on where the audience actually left each piece, read against the plan's own acts.
Performance
What stays yours
The same three rules across every template. They are the blueprint's, not the agents' — an agent cannot widen them.
Nothing publishes itself
Every tool that acts outside your organization — a post, an email, a reply — is granted
ask, notallow. The call parks the session with its arguments in front of you and waits. Approval is a person, every time.A budget it cannot raise
A daily cap per agent and a cap per task, declared in the template and enforced by the platform. A run that would exceed its cap stops rather than spends.
Denied by default
An agent is offered the tools its template names and nothing else. Everything it was not granted is denied by name — the shell included, which is also why a content agent never provisions a machine it would not use.
What it ships
A template is a repo: the apps below, their server, and a naive.config.ts declaring the agent team. Clone it, bind it to your organization, bring it up.
Naive Long Form v1 has no released repository behind it yet, so there is no build to describe. The apps it intends to deploy are channel.
- $ naive template media --template longform# clone the blueprint repo
- $ naive claim --key sk_...# bind the clone to your organization
- $ naive up# provision the apps, the agents, their budgets
The number
A target while the benchmark runs — an average across a cohort of new accounts, never revenue and never per-client.
1:12
avg. view duration at 90 days · 20 new channels
It gets a measured number once it runs on bench terms: live accounts, a fixed budget, twelve weeks, a control arm.
Launch your naive long form v1
An agent team, a workspace, a budget. No hiring.