
Security News
Agent Baseline Publishes Security Reference Architecture for Software Factories
Agent Baseline’s new reference architecture shows how to secure software factories with four boundaries enforced outside the AI agent.
writing-skills
Advanced tools
Procedural writing-style skills for agents — measurable standards, voice craft, and character voices as SKILL.md instructions.
16 writing-style skills for AI agents: 9 measurable frameworks, 7 character voices.
Plain SKILL.md files. Works in any agent that reads skills
(Claude Code, opencode, Codex, Cursor, Kiro, Gemini).
npx skills add writing-skills
One command, every agent on your machine.
| Harness | Command |
|---|---|
| Claude Code plugin | /plugin marketplace add surendranb/writing-skills then /plugin install writing-skills@writing-skills |
| Codex ≥ 0.147 / Kiro | git clone https://github.com/surendranb/writing-skills.git — plugin.json is auto-discovered |
| npm | npm install writing-skills |
| Python | uv add writing-skills or pip install writing-skills |
Ask for the style. The skill loads itself.
"rewrite this in plain language" · "make this a BLUF business update" · "press release in AP style" · "say it like Yoda"
Every skill ends with a Verify checklist the agent must pass before it delivers.
| You say | Skill that fires | What you get |
|---|---|---|
| "Simplify this insurance letter so anyone can read it" | plain-language | Sentences ≤25 words, passive voice stripped, jargon replaced — checked against the Plain Writing Act rules |
| "Turn these notes into an update for my VP" | business-writing | BLUF first line, the ask and deadline explicit, buzzwords deleted |
| "Write the release notes for v2.0" | google-dev-docs | Active voice, present tense, verb-first headings |
| "Draft the outage announcement" | corporate-communication | News in line one, reader impact named, zero euphemism |
| "Make this maintenance manual usable by non-native speakers" | asd-ste100 | Simplified Technical English: one instruction per sentence, no idioms |
| "Press release for the funding round, AP style" | journalism-ap | AP Stylebook attribution and number rules, no editorializing |
| "Explain recursion like Bob Ross" | bob-ross | A happy little explanation — calm, encouraging, and technically correct |
| "Rejection email, but make it Paddington" | paddington | Firm news delivered with unfailing politeness |
| Skill | Enforces |
|---|---|
orwells-rules | Orwell's 6 rules. Zero dead metaphors, short Saxon words, prune verbal limbs, active voice, barbarism override. |
plain-language | Plain Writing Act. Sentences ≤25 words, passive ≤10%, jargon ban list. |
business-writing | BLUF first line, explicit ask + deadline, buzzword ban list. |
corporate-communication | News first, reader impact named, zero euphemism. |
gov-uk-style | GOV.UK guide. Keyword-first headings, ≤5-sentence paragraphs. |
asd-ste100 | Simplified Technical English. One instruction per sentence, no idioms. |
google-dev-docs | Google style. Active voice, present tense, verb-first headings. |
journalism-ap | AP Stylebook. Attribution, number rules, zero editorializing. |
scott-adams | Simplicity-is-persuasion. Curiosity-hook openers, ≤15-word sentences, prune "very/really". |
ted-lasso · jack-sparrow · shrek · yoda · winnie-the-pooh · paddington · bob-ross
Copy template/SKILL.md, fill it, run
python3 scripts/validate_skills.py. CI enforces the contract — see
CONTRIBUTING.md.
MIT — Surendran B
FAQs
Procedural writing-style skills for agents — measurable standards, voice craft, and character voices as SKILL.md instructions.
We found that writing-skills demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.

Security News
Agent Baseline’s new reference architecture shows how to secure software factories with four boundaries enforced outside the AI agent.

Security News
arXiv now limits authors to two submissions a month as AI slop overwhelms moderators, delays good papers, and sparks debate over applying the limit to everyone.

Research
/Security News
A new GhostAction wave hits hundreds of GitHub repos, expanding CI/CD secret theft to cloud and AI credentials in source code and git history.