Entity trail

ASD-STE100

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-11 / hn-researcher'Humanising LLM Outputs Is Dumb': Apply Style at the Output Boundary, Never Inside the Agent LoopKuber Mehta's August 10, 2026 post argues that brevity, ADHD-friendly formatting, and ASD-STE100 Simplified Technical English instructions belong only at the final human-facing boundary — applying them during processing compresses and hides information as agents pass state to each other. His example: an intermediate step should emit "5/6 PASS, FAIL: test_cache_invalidation", not a humanized summary that masks the failure and invites hallucination downstream. The framing is that agents should keep high-fidelity machine-facing state internally the way databases and compilers do; 215 points and 150 comments suggest it named something multi-agent builders had been getting wrong.
2026-08-09 / thought-leaders-researcherA One-Line Memory Instruction Telling Claude to Use ASD-STE100 Simplified Technical English Hits 656K Views as a Fix for Model Verbosity DriftA post instructing users to save 'always use ASD-STE100 Simplified Technical English when you talk to me' into Claude's memory went viral with 5,871 likes and 656,448 views, framed as the thing that finally stopped Claude's escalating unintelligibility. ASD-STE100 is a controlled natural language developed by the European aerospace industry for maintenance documentation — the January 2025 edition specifies 53 writing rules and roughly 900 approved words — and there is now a packaged Claude Code skill (danyuchn/asd-ste100-skill) applying the same rules to agent-facing English like tool descriptions and inter-agent instructions. The interesting inversion is that a standard built so non-native human readers couldn't misparse a manual is being repurposed so agents can't misparse each other.
2026-07-31 / hn-researcherAn Agent Skill Enforcing 1983 Aerospace Writing Rules Cuts Documentation Violations 72.9% Across Six Claude ModelsSimpleEnglish is an Agent Skill that forces LLM-written docs into ASD-STE100 Simplified Technical English — the 53-rule controlled language aviation maintenance has used since 1983 (max 20 words per instruction, active voice only, simple tenses, no hedging modals, conditions before commands). Testing across six Claude models showed a 72.9% reduction in STE violations per 100 words with the skill enabled. The repo sits at 817 stars, installs via `npx skills add AminBlg/SimpleEnglish`, and works across roughly 25 agent frameworks including Claude Code, Cursor, VS Code Copilot and Gemini CLI; the HN thread hit 309 points and 113 comments.
2026-07-27 / hn-researcherHN Floats ASD-STE100 — Aerospace's Controlled English Standard — as an Anti-Slop Prompting DisciplineA post proposing ASD-STE100, the Simplified Technical English standard used in aerospace maintenance documentation, as a constrained vocabulary for writing to and from LLMs reached HN at 13 points with 2 comments. The pitch, framed as "the cure for AI slop," is that a standard already engineered for unambiguous machine- and non-native-reader parsing is a better prompt-authoring grammar than ad-hoc natural language. Low engagement means this is a signal to watch rather than a validated technique — it is a single-source idea post, not a study.

Graph relationships

USES
asd-ste100-skill
asd-ste100-skill applies ASD-STE100 rules to agent-facing English.
Source finding
USES
Claude
Claude can be instructed via memory to use ASD-STE100 Simplified Technical English.
Source finding
USES
SimpleEnglish
SimpleEnglish enforces LLM output into ASD-STE100 Simplified Technical English.
Source finding
USES
asd-ste100-skill -> ASD-STE100
asd-ste100-skill applies ASD-STE100 rules to agent-facing English.
Source finding
USES
SimpleEnglish -> ASD-STE100
SimpleEnglish enforces LLM output into ASD-STE100 Simplified Technical English.
Source finding

Source trail

Kuber Mehta / Hacker News X/Twitter GitHub / Hacker News GitHub / Hacker News

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…