agentleFS
zalando-incubatorzalando-incubator/agentic-identity-broker/.agents/skills/simple-english/SKILL.md

simple-english

Skill79 starsChanged 14 days agoMITView on GitHub

Write or rewrite technical text with the rules of ASD-STE100 Simplified Technical English so it is clear, unambiguous, and free of AI slop. Use for documentation, READMEs, runbooks, procedures, error messages, release notes, incident reports, and API guides. Also use when the user says "STE", "Simplified Technical English", "ASD-STE100", "de-slop", "make this readable", "write for non-native readers", or asks for docs that translate well. Enforces the standard's 53 rules: 20/25-word sentence limits, one word one meaning, simple tenses, active voice, condition before command.

27 other projects use a version of this, 20 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use xsai, pnpm and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • alint-devSkill

    xsai

    Use this skill when the user is building with `xsai` or any `@xsai/*` package, or is evaluating xsAI for a small OpenAI-compatible workflow with text generation, streaming, tool calling, structured output, embeddings, image generation, speech synthesis, or transcription.

    Also in 4 projects that use this

  • JetBrainsSkill

    pnpm

    Node.js package manager with strict dependency resolution. Use when running pnpm specific commands, configuring workspaces, or managing dependencies with catalogs, patches, or overrides.

    Also in 3 projects that use this

  • PydanticSkill

    i-have-adhd

    Shape every response for an ADHD reader who reviews the work but does not do the coding — the agent does. Use whenever responding to ANY message — coding, debugging, planning, casual. Lead with the result or the decision. Surface anything needing the user's input as a structured question, never buried in prose they'll skim past. No preamble, no recap, no closers. State errors matter-of-factly. Make finished work visible. Trigger even on casual messages and even when brevity wasn't requested.

    Also in 5 projects that use this

  • AWSSkill

    wait-what

    Stop. That last message did not land: re-pitch it.

    Also in 4 projects that use this

  • AWSSkill

    wizard

    Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.

    Also in 4 projects that use this

  • GoogleSkill

    setup-matt-pocock-skills

    Configure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.

    Also in 4 projects that use this

  • +2 more from aws

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. 0 are common, 10 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 25 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write for that tired reader.

    line 23
  2. Distinctive:

    When asked to write or rewrite technical text:

    line 27
  3. Distinctive:

    Do the self-check: before you deliver. This step is not optional.

    line 33
  4. Distinctive:

    Do not mix the two in one passage.

    line 54
  5. Distinctive:

    Wrong shortening: Ensure file exists before running.

    line 132
  6. Distinctive:

    STE: Make sure that the file exists before you run the command.

    line 133
  7. Distinctive:

    No imperative in descriptive text.

    line 159
  8. Distinctive:

    Never bury the instruction after the explanation.

    line 169
  9. Very specific:

    Search your draft for: 'll, 're, 's (contraction), has been, have been, should, shall, however, therefore, -ing verbs after a comma, …

    line 296
  10. Very specific:

    Search for every if and when.

    line 297
  11. Very specific:

    Search for the verbs you did NOT pick in Your Task step 3 (check, verify, confirm, ensure).

    line 298
  12. Distinctive:

    Check each vertical list: colon on the lead-in, items start with an uppercase letter, no comma or semicolon at the end of an item, no …

    line 299
  13. Distinctive:

    No tool can guarantee STE compliance.

    line 323

“Common” is measured: kept by at least 6 in 10 of the 25 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Simple English: Write Like an Aerospace Manual

Write technical text with the rules of ASD-STE100 Simplified Technical English. STE is the controlled language that aerospace and defense manufacturers use for maintenance documentation. The rules exist so that a tired reader who is not a native English speaker cannot misread an instruction. They remove the usual signs of AI-generated text as a side effect: long sentences, synonym rotation, hedges, filler, and decorative clauses.

Write for that tired reader. Each sentence must survive one read.

Your Task

When asked to write or rewrite technical text:

  1. Select the mode (pragmatic or strict — the table that follows).
  2. Classify each passage as procedural or descriptive. Every other rule depends on this.
  3. Correct your vocabulary before drafting. In strict mode, the dictionary rejects check/verify/confirm/ensure as verbs. Replace them by intent: make sure that (verify a state), examine (look for faults), measure (get a value). In pragmatic mode, pick one and keep it. Pick ONE noun for config/settings (all are valid technical nouns — pick one and keep it). Use no other word for these concepts in the whole document.
  4. Apply the rules from the catalog that follows.
  5. Do the self-check before you deliver. This step is not optional.
  6. Never touch code, identifiers, commands, or quoted errors (see Untouchables).

When asked to CHECK text instead of writing it, report each violation as: rule number, the offending text, a compliant rewrite. Cite only rule numbers that exist in this file. Do not cite rule numbers from memory. The numbering is unintuitive and invented rule numbers are a known failure.

Two Modes

ModeWhenWhat you apply
Pragmatic (default)Docs, READMEs, error messages — the user wants clear textAll structural rules. Domain words stay ("idempotent", "webhook").
StrictThe user names STE, ASD-STE100, or complianceStructural rules + full vocabulary discipline, and tell the user that full compliance needs the official dictionary (free at asd-ste100.org).

Step 1: Classify the Text

Procedural (instructions)Descriptive (explanations)
PurposeTell the reader what to doExplain what a thing is or does
Verb formImperative: "Install the pump."Simple present/past/future
Sentence limit20 words (Rule 5.1)25 words (Rule 6.3)
Unit ruleOne instruction per sentence (5.2)One topic per paragraph (6.5), max six sentences per paragraph (6.6)

Do not mix the two in one passage. A "Getting started" section is procedural. An "Architecture" section is descriptive. A note inside a procedure is descriptive (25-word limit, no imperative).

THE RULE CATALOG

53 rules in 9 sections, paraphrased from ASD-STE100 Issue 9 with software examples. The official wording is in the free standard at asd-ste100.org.

Section 1 — Words (Rules 1.1-1.14)

RuleInstruction
1.1Use only approved words, technical nouns, or technical verbs.
1.2Use an approved word only as its listed part of speech.
1.3Use an approved word only with its approved meaning.
1.4Use only the approved forms of verbs and adjectives.
1.5You can use domain words as technical nouns ("webhook", "commit", "endpoint").
1.6Use an unapproved word only when it is a technical noun or part of one.
1.7Do not use technical nouns as verbs.
1.8Use the technical nouns of your project or industry.
1.9When you pick a technical noun, pick a short and clear one.
1.10No regional, slang, or jargon words as technical nouns.
1.11One item, one name. Do not call it "config" here and "settings" there.
1.12You can use domain verbs as technical verbs ("deploy", "compile", "merge"). The standard names computer verbs as legal, for example: click, press, enter, type, tap, copy, cut, paste, delete, save, scroll, sort, validate, boot, debug, download, install, load, process, reboot, update, upgrade, upload. When a dictionary verb does the same job, prefer it: "find" instead of "detect".
1.13Do not use technical verbs as nouns.
1.14Use American English spelling.

In pragmatic mode, rules 1.5, 1.8, and 1.12 do the heavy lifting: your domain vocabulary is legal. The ones agents break are 1.7, 1.11, and 1.13.

Before: You can webhook the event, then do a deploy. After: Send the event to the webhook. Then deploy the service.

Section 2 — Multi-word nouns (Rules 2.1-2.2)

RuleInstruction
2.1Write multi-word nouns of three words or fewer.
2.2When a technical noun needs more than three words, write it in full once, then give a short form or hyphenate the units.

Break long noun chains with prepositions (of, on, in, for):

Before: the connection pool timeout configuration value After: the timeout value for the connection pool

Section 3 — Verbs (Rules 3.1-3.7)

RuleInstruction
3.1Use only the verb forms that the dictionary gives.
3.2Use only: infinitive, imperative, simple present, simple past, simple future, past participle as adjective.
3.3Use the past participle only as an adjective ("the cached response").
3.4No auxiliary verbs for complex constructions. No present perfect, no "is to be installed".
3.5Use an "-ing" form only as a technical noun or inside one ("logging", "the mounting bracket") — never as a verb.
3.6Active voice. In descriptive text, passive is legal only when the agent is unknown. To repair an agentless passive, use "you" (the reader) or "we" (your company) as the subject: "Indexes are not used on this table" → "We do not use indexes on this table."
3.7Describe an action with a verb, not a noun ("compress the file", not "perform compression of the file").

Approved modals: can, will, must. Banned: should, would, may, might, could. The standard rejects "could" even for possibility: write "an explosion can occur", never "could occur". For "should": a requirement becomes "must". A suggestion is stated as fact or deleted. This matters double for agent instructions — models read "should" as optional.

Before: The migration has completed and the table is being rebuilt. After: The migration completed. The database rebuilds the table.

Before: The flag can be set in the config file, making restarts unnecessary. After: You can set the flag in the config file. Then a restart is not necessary.

Before: The temperature must be adjusted. After: Adjust the temperature.

Section 4 — Sentences (Rules 4.1-4.5)

RuleInstruction
4.1Write short and clear sentences.
4.2Do not omit words or use contractions to shorten sentences. Keep articles, keep "that".
4.3Use a vertical list for complex text. Put a colon at the end of the lead-in. Start each item with an uppercase letter. An item gets a period only if it is a full sentence — never a comma or a semicolon. The last item gets a period. Do not mix instructions and facts in one list. Do not nest lists.
4.4Use connecting words between sentences on related topics ("Then", "As a result").
4.5Put an article (the, a, an) or a demonstrative adjective (this, these) before nouns where applicable. Exception: no article before a noun when an identifier follows it — "Restart pod web-7f9b2", not "Restart the pod web-7f9b2".

Rule 4.2 is the anti-terseness rule. STE is short sentences with complete grammar, not telegraph style:

Wrong shortening: Ensure file exists before running. STE: Make sure that the file exists before you run the command.

Section 5 — Procedural writing (Rules 5.1-5.5)

RuleInstruction
5.1Maximum 20 words per sentence. Warnings and cautions included.
5.2One instruction per sentence, unless two actions happen at the same time. A step can have a second sentence for an immediate result or limit: "Run the migration. The migration must take less than 5 minutes."
5.3Write instructions in the imperative: "Run the migration."
5.4Put a required condition before the command, divided by a comma: "If the build fails, read the log."
5.5Notes give information, never instructions, requirements, or limits. A limit belongs with its action in the work step. Notes get the 25-word limit. Notes-test: the procedure must still work for a reader who deletes all notes.

Before: You'll want to grab the API key from the dashboard before configuring the client, which you can do under Settings. After: Get the API key from the dashboard, under Settings. Then configure the client with this key.

Section 6 — Descriptive writing (Rules 6.1-6.6)

RuleInstruction
6.1Give information gradually: one new fact per sentence.
6.2Use key words and phrases to give the text a logical structure.
6.3Maximum 25 words per sentence.
6.4Group related information in paragraphs.
6.5One topic per paragraph.
6.6Maximum six sentences per paragraph.

No imperative in descriptive text. Descriptions explain. Procedures instruct.

Section 7 — Safety instructions (Rules 7.1-7.3)

RuleInstruction
7.1Use a word that shows the risk level ("WARNING" = injury, "CAUTION" = damage). If the two risks occur together, use "WARNING".
7.2Start with a clear command or condition.
7.3Then give the risk or the possible result.

Never bury the instruction after the explanation. The pattern transfers directly to destructive CLI flags, irreversible migrations, and dangerous API options.

Before: Note that data loss may occur in some circumstances if the destructive flag happens to be enabled when running against production. After: CAUTION: Do not use the --force flag against production. The flag deletes rows that do not match the source.

Section 8 — Punctuation and word count (Rules 8.1-8.7)

RuleInstruction
8.1All standard punctuation is legal except the semicolon. Write two sentences instead.
8.2Use hyphens to connect words that act as one unit.
8.3Parentheses are legal for references, item numbers, abbreviations, plural forms, explanations, alternatives.
8.4In a vertical list, the lead-in colon ends a sentence for word count. Each item after the colon counts as a new sentence and gets its own 20/25-word budget.
8.5Text inside parentheses counts as one word.
8.6Count as one word each: numbers, numbers with units, abbreviations, alphanumeric identifiers, quoted text, titles, labels, proper nouns.
8.7A hyphenated word counts as one word.

Rule 8.6 matters for software text: sqlpipe run --config sqlpipe.yaml in backticks is quoted text and counts as one word. Long identifiers do not blow your sentence budget.

Section 9 — Writing practices (Rules 9.1-9.4, GR-1 to GR-8)

RuleInstruction
9.1When a word-for-word replacement does not work, restructure the sentence.
9.2Use each approved word correctly: approved meaning, approved part of speech.
9.3Do not build phrasal verbs ("go down" → "decrease", "set up" → "install" or "configure").
9.4Keep one consistent style and terminology through the whole document.

General recommendations GR-1 to GR-8: keep the conjunction "that", be careful with "with", give pronouns clear referents, prefer "this + noun" over bare "this", avoid false friends, avoid Latin abbreviations, use inclusive language, and use the possessive apostrophe form only when you are sure it is correct (GR-8: if unsure, do not use it — non-native readers find it hard). GR-2 also kills a common habit: keep the primary verb first and the tool after "with" — "Fetch the URL with curl", not "Use curl to fetch the URL".

GR-6 for software docs: "e.g." → "for example", "i.e." → "that is", and delete "etc." — name the items or write "and more".

VOCABULARY DISCIPLINE

The official dictionary (~900 approved words, ~1,200 banned words with alternatives) is copyrighted by ASD and is not reproduced here. Its mechanics apply without it: one word, one meaning, one part of speech. Issue 9 adds a quick-reference list of approved verbs in the dictionary introduction — in strict mode, check your verbs against the official standard.

Known part-of-speech rulings, useful as patterns:

WordRuling
test, check, workNoun only. "Do a test", not "test the pump". "Check that X" becomes "make sure that X".
oilTechnical noun (TN) only. For the verb, the dictionary gives "lubricate": "Lubricate the linkage with oil."
helpVerb only. For the noun, the dictionary gives "aid": "with the aid of".
fall (noun)Rejected. Use "decrease" for a reduction in value. Use FALL (verb) only for physical movement downward by gravity: "Make sure that the tools do not fall into the engine."
follow"To come after" only, never "obey". Write "obey the instructions".
above, belowPhysical positions only. For limits write "more than", "less than".

The modal ladder

You wroteSTE writes
should (requirement)must
should (recommendation)Delete it, or state it as fact: "X is better because Y."
should (inverted conditional: "should a failure occur")if: "If a failure occurs"
may / might / could (possibility)can
may (permission)can
would (hypothetical)can, or restructure: "If X occurs, Y occurs."

Slop-to-simple substitutions

AI-generated docs overuse a known set of words. references/word-swaps.md maps each one to a plain replacement. Read it when you rewrite existing text. If a word carries no fact, delete it instead of replacing it.

Consistency pass

Collapse synonym rotations to one term each (Rules 1.11, 9.4). The two lists that follow work differently.

Technical nouns — not in the dictionary. Pick one and keep it consistent (both modes):

  • config / configuration / settings / options → pick one

Dictionary rulings — the standard has already chosen. Use the approved word (strict mode). Or pick one and keep it consistent (pragmatic mode):

You wroteDictionary statusUse instead
check (verb) / verify / confirm / ensureAll rejected as verbsStrict, by intent: make sure that (verify a state), examine (look for faults: "examine the log"), measure (get a value), or the noun: "do a check of". Pragmatic: pick one and keep it.
validateNot in dictionaryUse as technical verb (Rule 1.12), or replace with make sure that
delete / drop (verb) / destroyAll rejected as dictionary verbsStrict: erase (data), remove (physical). In computer contexts delete is also a legal technical verb (Rule 1.12). Avoid drop and destroy.
removeApproved verbKeep it
run / executeBoth rejectedoperate for run, do for execute (strict); pick one (pragmatic)
invoke / launchNot in dictionaryUse as technical verbs (Rule 1.12)
display (verb) / render / present (verb)All rejectedshow covers most software cases. Official alternatives: display → show, render → make, present → give, show
issueNot in dictionaryUse as technical noun, or replace with problem (approved)
failureRejected in general use; approved as TN for performance lossUse only when it means a performance error: "a failure of the pump"
errorApproved nounKeep it
problemApproved nounKeep it

Recurring errors the standard names

The dictionary introduction lists the words that writers get wrong most often. This is the software-relevant set, given as rulings only — the full list is part of the official dictionary.

You wroteSTE writes
howeverbut
thereforethus, as a result
since (= because)because
anyDelete it, or restructure: "if you have any questions" → "if you have questions"
nowat this time — better, delete it: "now start the service" → "start the service"
need to, have toImperative in procedures ("install"); "it is necessary to" in descriptive text
performdo
insertput (but SQL INSERT stays: it is quoted text)
reachget, get to
avoidprevent
repeatdo … again
acceptablepermitted — better, give the limit: "a latency of less than 200 ms"
complete (adjective)completed
the example below, the section aboveName the target, or put the reference after it: "the example that follows"

Untouchables

These are technical names (Rules 1.5, 8.6). Leave them exact, even when they break vocabulary rules:

  • Code blocks, inline code, identifiers, CLI commands, flags, file paths
  • Quoted error messages and log lines
  • Product names, API endpoint names, config keys
  • UI labels and button names ("click the Save button" — quoted text, counts as one word)
  • Numbers with units — each counts as one word in the sentence limit

Facts are untouchable too. Rewrite the style, not the content. When the source does not give a number, a cause, or an exact term, keep the general statement. Do not invent specifics to look concrete.

Beyond Documentation

The same rules apply to error messages, runbooks, incident reports, release notes, commit messages, agent instructions, support macros, UI copy, and translation prep. Read references/use-cases.md when the task is one of these. It gives the pattern for each.

Self-Check Before You Deliver

This step is not optional. Run these five checks on your draft:

  1. Count words in your three longest sentences. Over the 20/25 limit → split them.
  2. Search your draft for: 'll, 're, 's (contraction), has been, have been, should, shall, however, therefore, -ing verbs after a comma, semicolons.
  3. Search for every if and when. Each one stands at the START of its sentence, before the command. "Increase the timeout if the network is slow" → "If the network is slow, increase the timeout."
  4. Search for the verbs you did NOT pick in Your Task step 3 (check, verify, confirm, ensure). STRICT MODE: route each hit by intent — make sure that, examine, or measure. Pragmatic mode: replace each hit with your chosen verb.
  5. Check each vertical list: colon on the lead-in, items start with an uppercase letter, no comma or semicolon at the end of an item, no procedural and descriptive items mixed.

Fix what you find, then deliver. For a full audit, run references/checklist.md.

Full Example

Before (real unedited AI output):

Connection timeouts. If sqlpipe hangs or fails with dial tcp: i/o timeout, check that the host running sqlpipe can reach the Postgres port (usually 5432) — this is often a security group or firewall rule blocking the connection. If you're connecting to a managed database (RDS, Cloud SQL, etc.), confirm the instance allows connections from sqlpipe's IP. You can also try increasing source.connect_timeout_seconds in your config, since a slow network path can trip the default timeout even when the connection eventually succeeds.

After (classified procedural, verb = "make sure", conditions first, one instruction per sentence):

Connection timeouts. sqlpipe stops with dial tcp: i/o timeout when it cannot connect to the Postgres port (5432 by default).

  1. Make sure that the host that runs sqlpipe can connect to the Postgres port. A firewall or security group usually blocks it.
  2. If the database is managed (RDS, Cloud SQL), make sure that the instance accepts connections from the IP of sqlpipe.
  3. If the network is slow, increase source.connect_timeout_seconds in the configuration.

What changed: 40-word sentences split under 20. "you're" expanded. "check/confirm" collapsed to "make sure that". Every condition moved before its command. "etc." removed. Code and error strings untouched.

Limits

STE is for technical facts and instructions. Do not apply it to marketing copy, blog voice, or brand writing — it deletes persuasion by design. When a user asks for STE on marketing text, say so and offer it for the docs instead.

No tool can guarantee STE compliance. If the user asks for a compliance claim, say that.

References

  • references/checklist.md — full verification pass with searchable patterns, for check mode and final audits
  • references/word-swaps.md — slop-to-simple word map, for rewriting existing text
  • references/use-cases.md — long-form adaptations: error messages, runbooks, incident reports, commits, UI copy, i18n

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.