Skills
Structured blog post production
from agent notes to editorial voice
anycap-blog-production is an installable skill that teaches Claude Code, Cursor, Codex, and similar coding agents how to produce structured, publication-ready blog posts from raw input. Without this skill, an agent generating a blog post cannot look up current statistics, verify product claims against live sources, or maintain a consistent editorial structure across long outputs. With the skill installed, the agent gains access to a production workflow that starts from user-provided notes or briefs, enriches them with web search and web crawl data for first-party evidence, and formats the result into structured sections with a consistent AnyCap editorial voice. The workflow also supports image generation for featured visuals, making it useful for teams producing content pipelines where both the text and the visual assets need to come out of the same agent session. This is particularly valuable for developer marketing teams, content operations, and any team that wants repeatable blog production quality without manual copy editing after every run. The skill installs through skills.sh or the AnyCap CLI.
Install time
< 5 min
Supported agents
Claude Code · Cursor · Codex
Platform
macOS · Linux · Windows
Install this skill
npx -y skills add anycap-ai/anycap -s 'anycap-blog-production' -g -y
How the workflow changes after install
BEFORE
Agent writes a blog post from training data alone. No live sources, no citations, no consistent structure. Output quality varies per run.
AFTER
Agent starts from your notes, searches for current evidence, structures the post in a defined editorial format, and delivers a publication-ready draft with citations.
- 1
Receive brief
User provides notes, a topic, product facts, or a rough outline. Brief quality directly affects output quality — a paragraph of context produces a significantly better post than a single topic phrase. Include target audience, publication, and tone guidelines where available.
- 2
Gather live evidence
Agent calls `anycap web search` to find current statistics, recent publications, and reference sources that support the post's claims. The skill runs 3–5 targeted queries covering different evidence angles — statistics, case studies, and expert opinions — to avoid building the post on a single narrow source.
- 3
Crawl reference pages
Agent calls `anycap web crawl` on the 2–4 most authoritative reference URLs to pull full article content. This step surfaces specific numbers, methodology details, and quotable passages that live inside articles but never appear in search snippets.
- 4
Structure the post
Agent assembles content into H2/H3 sections following AnyCap's editorial template: intro with hook, 3–5 body sections with evidence blocks, and a conclusion with CTA. Each claim is attributed to a crawled source. The post ships as structured Markdown, ready for CMS import or review.
- 5
Generate featured image (optional)
If configured, agent calls `anycap image generate` to produce a featured visual aligned with the post topic, with dimensions pre-set for the target publication — 1200×630 for blog header, 1200×675 for social sharing preview.
SAMPLE OUTPUT
Blog post complete: "Why AI agents need structured memory" ├── Evidence gathered: 5 sources (live, 2026-04-18) ├── Word count: 1,920 ├── Citations: 5 inline ├── Sections: intro + 4 body + conclusion └── Featured image: blog-hero.png (1200×630)
Capabilities used by this skill
Supported agents
Frequently asked questions
- What does this skill do that direct prompting doesn't?
- Direct prompting writes from training data. This skill adds live evidence gathering, consistent editorial structure, and optional image generation — turning Claude into a repeatable content production pipeline.
- What format does the output use?
- Structured markdown with H2/H3 headings, evidence blocks with citations, and a consistent editorial voice. You can provide a custom format spec in the brief.
- Can it generate the featured image too?
- Yes. The skill includes an optional image generation step using AnyCap image generation.
- Do I need an AnyCap account?
- Yes. Install the CLI and run `anycap login` before your first session.