Works with the AI Agents you already use
Research-Backed Article Writer (Cited, On-Topic, WordPress-Ready)
Produce long-form articles that are researched, cited and on-topic: staged pipeline, source credibility scoring, evidence-pack drafting, automatic checks for unsourced numbers and topic drift, and WordPress publishing.
$15
Research-Backed Article Writer (Cited, On-Topic, WordPress-Ready)
Example session with this skill installed
Our marketing team publishes AI-written blog posts on WordPress, but readers found made-up statistics and some posts wander off the topic we gave. Redesign the pipeline in PIPELINE.md so this stops, and say what gets checked before anything is published.
- Read your context and instructions
- Compiled the research-backed article writer
PIPELINE.md: analysis → research plan → grounded findings with URL and year → source scoring (drop below 0.65) → evidence pack [1..n] → draft that cites every number → article_check (unsourced numbers, dangling citations, topic coverage ≥ 0.6, structure, filler) + judge (coverage ≥ 0.7) → one repair → WordPress draft via an Author account's application password. The topic string is passed unchanged to every stage and used as the title. Each article logs cost, tokens, sources and quality score.
Connects securely to your tools. The creator never sees your data.
About this skill
Single-prompt article generation fails in three predictable ways: it drifts off the topic it was given, it states numbers no source supports, and it reads like every other AI article. This skill gives your agent a staged pipeline where research happens before writing, the writer only gets an evidence pack of scored sources, and code checks the result before anything is published.
What your agent builds with it
- Stage contracts with structured output: topic analysis, audience and narrative, research plan, grounded findings, evidence pack, draft, verification. The original topic is passed verbatim through every stage (the main cure for drift).
- Source credibility scoring (
source_score.py): authority by domain and publication type, currency, verifiability and relevance; drops weak or untraceable sources. - Drafting rules: every number, percentage or money amount carries a citation to the evidence pack, or the sentence has no number.
- Pre-publish checks (
article_check.py): unsourced numbers, dangling and unused citations, topic coverage of title, intro and headings, structure, word count, filler phrases, repetition; plus an LLM judge for drift with one repair round. - WordPress publishing: application passwords for a non-admin author, media upload with alt text, sanitised HTML, categories and tags, drafts by default, idempotent updates.
- Cost and quality log per article.
Included: SKILL.md, pipeline reference (schemas and prompts), WordPress reference, two standard-library scripts, and an eval case.
Part of the Production AI Agents series. Pair with LinkedIn Content Autopilot to promote each article.
Checks catch unsourced numbers and drift, not every factual error in prose; a human review step is still recommended for high-stakes content. The domain authority table is a starting point to tune for your field. WordPress is covered in depth; other CMSs need their own publisher.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 day ago
Requires
Permissions
File Scopes
The bundled scripts score sources and check articles offline; they make no network calls.