writing-for-agents
Reference for writing documents an agent consumes: skills, AGENTS.md or CLAUDE.md, and docs reached by a pointer. Invoke when creating or editing a skill or an agent instruction file, or when a document's pointer wording, load, or structure is in question. Do not invoke merely because the user mentions writing-for-agents, names a skill, or writes 'skill <word>'.
- License 许可
- MIT
- 最近更新
- 2026-11-12
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- writing-for-agents
- description
- Reference for writing documents an agent consumes: skills, AGENTS.md or CLAUDE.md, and docs reached by a pointer. Invoke when creating or editing a skill or an agent instruction file, or when a document's pointer wording, load, or structure is in question. Do not invoke merely because the user mentions writing-for-agents, names a skill, or writes 'skill <word>'.
- disable-model-invocation
- true
02 资源文件 1
- 参考文档SKILL-MECHANICS.md2.6 KB
03 SKILL.md 正文
Writing for Agents
Contents
- Plain language - the four ISO 24495-1 principles and how the skill's levers reach them.
- Context pointers - descriptions and links that trigger reference loading.
- The two loads - agent context cost and human document recall.
- Information hierarchy - steps, reference, disclosure, co-location, and outlines.
- Steps and completion criteria - checkable, exhaustive stopping points.
- When to split - sequence and invocation boundaries.
- Leading words - shared concepts and positive instructions.
- Pruning - one source of truth, relevance, and removing no-ops.
Reference for writing any document an agent consumes: a skill, an AGENTS.md /
CLAUDE.md, or a doc reached by a pointer. The packaging differs; the writing
does not: the same levers make each one predictable, since the agent takes the
same process every run rather than producing the same output.
Write instructions and user-facing explanations in plain language. Keep precise terms, commands, and real limits; use code or mathematical notation only when the subject needs it, not to frame ordinary workflows.
When the document you're writing is a skill, read SKILL-MECHANICS.md for frontmatter, invocation choice, and router skills.
Plain language
Plain language is communication that puts the reader first. A document is plain when the reader can do four things on a first reading (ISO 24495-1:2023):
- Relevant: get the information they need, and only that.
- Findable: locate it without searching.
- Understandable: understand it on the first pass.
- Usable: act on it.
Treat the four as the outcome to check, not a style to apply. The skill's own levers already reach them: relevant is pruning, findable is context pointers, understandable is the information hierarchy, and usable is steps with checkable completion criteria.
Context pointers
A context pointer is a reference held in the agent's context that names some
out-of-context material and encodes the condition for reaching it. A skill's
description is one; a line in AGENTS.md naming a doc is the same object. The
pointer's wording, not its target, decides when the agent reaches the
material, and how reliably. A must-have target behind a weakly worded pointer is
a variance bug: sharpen the wording first, and inline the material only if
sharpening fails.
A pointer does two jobs: state what the material is, and list the branches that should trigger reaching it (a branch is a distinct case the document handles, so different runs take different paths through it). Every word of an always-loaded pointer costs on every turn, so it earns even harder pruning than the body:
- Front-load the leading word: the pointer is where it does its triggering work.
- One trigger per branch. Synonyms that rename a single branch are one branch written twice; collapse them and keep only genuinely distinct branches.
- Cut identity the body already carries.
The two loads
Every document and pointer you add spends one of two budgets:
- Context load is the cost of always-loaded material on the agent's window:
an
AGENTS.mdline, a skill description, anything sitting in context every turn, spending tokens and attention whether or not it fires. - Cognitive load is the cost on the human: which documents exist and when to reach for each. The human is the index. Not a cost to minimise: it is the price of human agency; spend it where human judgement matters, remove it where it does not.
Material reached only through a pointer escapes context load at the price of the pointer's own line; material with no pointer at all rides entirely on cognitive load.
Information hierarchy
A document is built from two content types: steps (the ordered actions the agent performs) and reference (definitions, rules, facts consulted on demand). The two mix freely: all steps (a recipe), all reference (a review's rules, this skill), or both. The core decision is where each piece sits on the information hierarchy, a ladder ranked by how immediately the agent needs the material:
- In-file step is the primary tier: what the agent does, in order.
- In-file reference is consulted on demand. Often a legitimately flat peer-set (every rule of a review on one rung), which is a fine arrangement, not a smell.
- Disclosed reference is pushed out into a separate file, reached by a context pointer, loaded only when the pointer fires. Spans a sibling file in the same folder through fully external reference that lives anywhere and any document can point at.
Push too little down and the top bloats; push too much and you hide material the agent actually needs. That tension is the whole decision.
Progressive disclosure is the move down the ladder (out of the main file and behind a pointer) so the top stays legible. Not primarily a token optimisation: it is how the hierarchy is protected. Branching is the cleanest disclosure test: inline what every branch needs, and push behind a pointer what only some branches reach. When a document has steps, in-file reference that should be disclosed buries them and turns attending to them into a coin-flip: a variance lever, not just a legibility one.
Co-location is the within-file companion: where the ladder decides how far down a piece sits, co-location decides what sits beside it once there. Keep a concept's definition, rules, and caveats under one heading rather than scattered, so reading one part brings its neighbours with it. The test: the document should read like documentation written for the agent. Grouped material reads that way; scattered material does not. (Distinct from duplication: that repeats one meaning in two places; scattering fragments one meaning across many.)
Sprawl is the failure mode here: a document simply too long, even when every line is live and unique. Attention thins across the excess, and every extra line is one more to keep relevant. The cure is the ladder: disclose reference behind pointers, and split by branch or sequence so each path carries only what it needs.
Outline: a document over 200 lines must open with a ## Contents section.
It is a bulleted list in document order, one item per top-level section,
formatted - **<heading>** - <the section's keywords and one-line description>,
so a read truncated at 200 lines still reveals what follows.
Steps and completion criteria
Every step ends on a completion criterion, the condition that tells the agent the work is done. Two properties make it a lever:
- Clarity: can the agent tell done from not-done? A vague bound ("understanding reached") invites premature completion: ending the step before it is genuinely done, attention slipping to being done. The visible steps still ahead (the post-completion steps) supply the pull; the criterion's clarity is the resistance. Defend in order: sharpen the bound first (local and cheap); only if it is irreducibly fuzzy and you observe the rush, hide the later steps by splitting the sequence. Hiding only works across a real context boundary (a hand-off or a subagent dispatch; an inline call leaves the later steps in context and clears nothing).
- Demand: how much it requires. "Every modified model accounted for" forces thorough work where "produce a change list" does not. Demand drives legwork (the digging the agent does within the work, latent in the wording rather than written as its own step), and it is not step-bound: "every rule applied" binds a body of flat reference just as "every step done" binds a sequence, which is how an all-reference document still carries an exhaustiveness bar.
The strongest criteria are both checkable and exhaustive.
When to split
Splitting one document into two spends one of the two loads, so split only when the cut earns it:
- By sequence: split a run of steps where the post-completion steps tempt the agent to rush the one in front of it. Keeping them out of view drives more legwork on the current task. Beware the reverse: merging sequences exposes each step's later steps to what follows, inviting premature completion.
- By invocation, skill-specific: see SKILL-MECHANICS.md.
Leading words
A leading word is a compact concept already living in the model's pretraining that the agent thinks with while running the document (lesson, fog of war, tracer bullets). Repeated as a token, never as a sentence, it accumulates a distributed definition and anchors a whole region of behaviour in the fewest tokens, by recruiting priors the model already holds. Coining your own works if you define it clearly, but a made-up word recruits no priors: you pay in definition tokens what a pretrained word gives free; reach for an existing word first.
It anchors twice. In the body, execution: the agent reaches for the same behaviour every time the word appears, and inside flat reference it focuses attention on a class of thing to look for. In a pointer, invocation: when the same word lives in your prompts, your docs, and your codebase, the agent links that shared language to the material and reaches it more reliably.
Hunt for opportunities to refactor with leading words. A triad spelled out at three sites, a pointer spending a sentence to gesture at one idea. Each is a passage begging to collapse into a single token:
- "fast, deterministic, low-overhead" becomes tight (a tight loop).
- "a loop you believe in" becomes red, turning a fuzzy gate into a binary observable state (the loop goes red on the bug, or it doesn't).
You win twice: fewer tokens, and a sharper hook for the agent to hang its thinking on. Assume every document is carrying restatements that leading words retire. Go find them.
Negation is the failure mode beside this lever: steering by prohibition drags the forbidden behaviour into context and makes it more available, not less. Don't think of an elephant, and the elephant is all there is; the negation is a weak modifier the strongly-activated concept overruns, so the ban half-reads as an instruction to do the thing. Prompt the positive: state the target behaviour ("write one-line comments") so the banned one is never spoken. A prohibition earns its place only as a hard guardrail you cannot phrase positively; even then, pair it with the positive target so attention lands on what to do.
Pruning
- Keep each meaning in a single source of truth: one authoritative place, so changing the behaviour is a one-place edit. Duplication (the same meaning in more than one place) costs maintenance and tokens, and inflates a meaning's prominence on the ladder past its real rank. (The accidental inverse of a leading word, which repeats a token on purpose, never the meaning.)
- The environment is a source of truth too (
package.jsonscripts, config files, the directory layout,--helpoutput), and a document that restates it is a cache: a copy of a lookup, earning its load only when the lookup is expensive. Cache what the agent cannot find by looking: the unwritten convention, the reason behind a choice, the gotcha no config confesses. Leave the one-file, one-command lookups to the environment, where they cannot go stale. - Check every line for relevance: does it still bear on what the document does? A line loses relevance by never bearing on the task (mere exposition, or a branch that should be disclosed) or by going stale as the behaviour or world it describes changes. Shorter documents are easier to keep relevant. Without a pruning discipline the default fate is sediment: stale layers that settle because adding feels safe and removing feels risky, until you must core down through them to find what is still live.
- Hunt no-ops sentence by sentence: an instruction the model already obeys by default pays load to say nothing. The test (does it change behaviour versus the default?) is model-relative, not reader-relative: two people disagreeing about a no-op disagree about the default, and settle it by running the document, not by debate. When a sentence fails, delete the whole sentence rather than trim words from it. The test also grades leading words: a word too weak to beat the default (be thorough when the agent is already thorough-ish) is a no-op, and the fix is a stronger word (relentless), not a different technique.
04 同仓库更多 Skills
wizard
Generate an interactive bash wizard for steps only a human can perform. Invoke when the user asks for a script or wizard for manual setup: infrastructure provisioning, credentials or CI secrets, a third-party dashboard, or a one-off migration or cutover. Do not invoke for steps the agent can perform, or merely because the user mentions wizard, names a skill, or writes 'skill <word>'.
wait-what
Stop and re-pitch the last message in plain, simplified language.
to-tickets
Split an agreed specification into reviewable vertical-slice issues with explicit blockers, then publish them after confirmation.
to-spec
Turn the current conversation into a specification and publish it after confirmation.
to-pr
Create a pull request from the current branch.
tdd
Test-driven development. Invoke only when the user asks to build a feature or fix a bug test-first, requests a red-green-refactor loop, or requests integration tests.