talocode-content-agent

Files (1)

• SKILL.md

SKILL.md (first lines)

# talocode-content-agent

Create original, evidence-led Talocode and TeraAI content through a repeatable brief-to-improvement loop.

## Purpose

Turn a specific audience problem into a useful, original social asset with visible proof, human approval, and a performance receipt that improves the next run.

## When to Use

Use this skill when:

- Planning a Talocode product Short, demo, post, or thread
- Planning useful TeraAI learning content
- Adapting a proven content pattern into an original Talocode artifact
- Reviewing a draft before publication
- Recording content performance and selecting the next experiment

## Core Loop

Use the leading word **receipt**: every stage leaves a reviewable artifact before the next stage begins.

```text
brief -> draft -> evidence review -> approval -> publish -> content factory reuse -> performance receipt -> improve next run
```

## Workflow

### 1. Brief

Write one brief for one audience and one problem.

```text
Audience:
Specific failure:
Useful outcome:
Visible proof:
One destination:
Channel lane:
```

Completion criterion: the brief names one real failure and one demonstration that can prove the proposed outcome.

### 2. Draft

Create the master script before adapting it for any platform.

For short-form video:

```text
0-3s: specific failure or result
3-12s: why the usual approach fails
12-45s: visible workflow and evidence
45-60s: outcome and one destination
```

Use natural language, readable captions, and original visuals. Show the product action rather than making broad capability claims.

Completion criterion: the draft has one clear hook, one proof sequence, and one destination.

### 3. Evidence Review

Verify the content before asking for approval.

- Check every technical, product, and performance claim against a real result.
- Confirm the destination works.
- Confirm visual proof matches the words on screen.
- Remove claims that cannot be demonstrated.
- Ensure external names, marks, assets, and references do not appear in public-facing content.

Completion criterion: every claim in the asset has visible proof or is removed.

### 4. Approval

Require human approval for public publishing, product claims, performance claims, paid promotion, and brand-sensitive replies.

Completion criterion: an approved review copy, title, caption, and destination exist before any public upload.

### 5. Publish

Keep content lanes separate.

- Talocode: developer failures, product workflows, build evidence, and open-tool infrastructure.
- TeraAI: learning, recall, revision, and study workflows until broader product access is available.

Publish one Talocode Short per day. Adapt the master asset for each destination without changing the supported claim.

Completion criterion: the published asset has the correct lane, destination, and platform-specific copy.

### 6. Performance Receipt

Review each public asset after at least 72 hours.

```text
Asset:
Date:
Hook:
Audience:
Views:
Average watch percentage:
Early retention:
Shares:
Comments:
Subscribers or followers gained:
Destination traffic:
What worked:
One next change:
```

Completion criterion: select one tested change for the next asset rather than changing several variables at once.

### 7. Content Factory

Treat every approved asset as raw material for several pieces, not one upload. One strong build can become a portfolio.

- **5x rule:** after the main asset publishes, adapt the same visible proof into up to five pieces (Short, full explainer, GitHub thread, docs entry, cross-post) without changing the supported claim. Different frame, same demonstration.
- **Reuse without repetition:** every derivative asset must preserve the approved claim and evidence, but change the hook, framing, depth, pacing, or audience angle. Do not publish near-identical captions/scripts across Short, explainer, thread, and docs.
- **Two tiers:** keep the daily problem-led Short lane plus an evergreen "what is / how" explainer layer (3-10 min) that can rank in search for months. Evergreen assets compound views the way a daily hit runway builds followers.
- **Decision metric:** when choosing the next experiment, prioritize the asset with the highest retention (average watch %) and shares per view, not the highest raw view count. Engaged retention predicts the next day's algorithm pick better than a single spike.

From talocode/skills, MIT.

More from talocode/skills