Imported from shipshitdev/skills (
skills/technical-writing/SKILL.md). Install upstream withnpx skills add shipshitdev/skills --skill technical-writing. Copyright stays with the author (MIT).
Technical writing
The goal is writing a tired engineer understands on the first read. Four layers: what kind of document this is, how sentences address the reader, how much each sentence carries, and whether any sentence reads two ways.
Companion to docs (repo-convention docs writer) and deslop
(prose-tell catalog). Apply
references/prose-slop.md from the selected deslop skill directory to every doc this skill
touches.
Authorized Scope
Apply this engine only within the user's requested task and existing explicit authorization. Loading or delegating to it grants no additional authority. Preserve report-only restrictions and the caller's target, host, provider, and cost limits. Existing approval satisfies a gate only for the same actions and scope; obtain approval before expanding them. Forward these limits to delegates.
Contract
Inputs:
- A doc, RFC, README, PR description, or commit message to write or review
Outputs:
- The rewritten surface, in one Diátaxis mode when the artifact is a document
Creates/Modifies:
- The named prose files or git metadata the caller asked for
External Side Effects:
- None beyond those writes
Confirmation Required:
- None for drafting. Confirm before committing.
Delegates To:
- File pointer:
references/prose-slop.mdfrom the selecteddeslopskill directory
Rules above the layers
- Cut every word that does no work.
- Use the short, everyday word.
- When a rule makes a sentence worse, fix the sentence another way.
- The codebase is the word list. Write the real symbol, file, flag, or command name.
Pick the mode first (Diátaxis)
- Action + learning: tutorial.
- Action + work: how-to.
- Understanding + work: reference.
- Understanding + learning: explanation.
Do not mix modes. Split and link instead. Source: diataxis.fr.
Tutorial. Open by saying what the learner will build. Every step produces a visible result. Write as "we", in commands.
How-to. Solve a problem a person has. Assume competence. Action only. Name the guide by the task.
Reference. Describe. Only describe. Mirror the structure of the thing described.
Explanation. One bounded topic. Anchor on a real why. Opinion is allowed here and nowhere else.
Sentences (Google developer style)
Talk to the reader as "you", present tense. Say who does what. Instructions as commands. Condition before the instruction. Common case first. Sentence-case headings. Numbered lists for sequences. Source: developers.google.com/style.
Load one thought at a time (STE)
One instruction per sentence. Split instructions longer than about 20 words. Keep "the" and "a". One word per action, then keep it. Source: asd-ste100.org (Issue 9, 2025).
Leave no sentence open to two readings (Global English)
Keep "only" and "not" next to the word they change. Break up long noun strings. Make every "it", "they", and "this" point at one thing. No slashes. Call each thing by one name. Source: Kohl, The Global English Style Guide.
Voice
- PR descriptions and commit messages are writing too. Every layer except Diátaxis applies to them.
- Product UI strings are not documentation.
- Write real paths and real symbols. Make every count true at the commit that lands it.
Review checklist
- Is each file one Diátaxis mode?
- Is every instruction a command, with its condition in front?
- Does any sentence carry two thoughts? Split it.
- Can any word be cut? Cut it.
- Is "only" next to the word it changes?
- Does each thing have exactly one name?
- Would a developer say these words out loud?
- Are symbols, paths, and counts real at this commit?
Technical Writing procedure
Read technical-writing procedure when running this workflow. Apply the authorized scope and mode of this entry point to every step. Resolve other skills through this distribution’s active catalog; resolve resources relative to the installed skill directory.