# Personal Blog

> Use when researching, outlining, drafting, revising, or polishing a personal blog article—an explainer, idea essay, personal essay, cultural review, or technical post—or when asked to preserve voice in a personal blog draft, turn notes into a personal blog post, 写博客, or 写一篇博客. Not for newsletters, magazine profiles, manuscript editing, email, chat, repository documentation, marketing copy, fiction, academic papers, or generic grammar cleanup.

- Skill: `misoto22/personal-blog` (Agent Skill, multi-file: 6 files)
- Install (CLI): `npx skillmds@latest add misoto22/personal-blog`
- Raw SKILL.md: https://api.skillmd.com/api/skills/misoto22/personal-blog/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- License: MIT
- Author: Misoto22 (https://skillmd.com/u/misoto22)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/misoto22/personal-blog

---


# Personal Blog

Build the requested blog artefact from evidence the writer supplied or sources
you verified. Preserve the writer's position and observable voice; never supply
the life, certainty, or results the source material does not contain.

## Workflow

1. **Resolve the artefact.** Extract the topic, intended reader, purpose, source
   material, language, and constraints already present. Produce only what was
   requested: research notes, an outline, editorial findings, a revision, or a
   finished article. Do not turn a limited request into a full drafting cycle.
   Read [shared/tone.md](shared/tone.md) and [shared/format.md](shared/format.md)
   before producing any artefact.
2. **Select one primary mode.** Read [Mode recipes](references/modes.md), choose
   `explainer`, `idea-essay`, `personal-essay`, `review`, or `technical`, and use
   that mode's structure and integrity check. For a genuine hybrid, borrow only
   the necessary checks from one secondary mode.
3. **Establish the voice basis.** Read [Voice and evidence](references/voice.md).
   Derive a task-local profile from the existing draft, representative samples,
   notes, or explicit directions. With no voice evidence, use a neutral
   provisional voice and make no personal claims.
4. **Separate the evidence.** Keep four distinct working categories: supplied
   personal material, externally verifiable claims, interpretation, and open
   gaps. Never turn an interpretation into a supplied fact.
5. **Verify what needs support.** Research externally checkable claims whose
   accuracy matters. Check any claim that could have changed against a current
   source. Prefer primary sources for technical or current claims, verify each
   citation before using it, state material versions and assumptions, and
   preserve disagreement between authoritative sources. Never invent an
   experience, memory, relationship, preference, belief, emotion, quotation,
   outcome, source, citation, command result, or test result.
6. **Choose earned structure.** Organize around the mode, evidence, and thesis.
   Do not impose a hook, three-part body, recap, or call to action.
7. **Draft or edit.** Preserve supplied facts, quotations, position, uncertainty,
   intentional roughness, repetition, fragments, and bilingual phrasing. If
   personal material is required but absent, remain impersonal or leave a visible
   bracketed Markdown gap for the writer.
8. **Run the mode-specific integrity check.** Remove unsupported claims and
   generic filler. Confirm that links resolve to the sources actually used and
   that no unobserved result is presented as fact.

## Delivery

Return the requested artefact without a process preface or delivery commentary.
For a finished article, return raw Markdown directly: no enclosing code fence,
no explanation outside the article, and no generic call to action. Use a
different format only when the user asks for one.