v0.20.0 — unified installer + launch-ready repo

TLDR v0.20.0

Verdict first. Filler never.

What's new

• Unified repo launch surface for the prompt + 30+ agent installer.
• GitHub Pages landing page enabled: https://jqbit.github.io/TLDR/
• Root clutter reduced: install/community/legal docs moved under docs/ and .github/.
• docs/LAUNCH.md added with ethical launch assets, copy, demo script, and channel plan.
• Package hygiene hardened: .npmignore, explicit package.json files, and bytecode exclusion checks.
• CI hardened: Node 20/22/24 matrix, Markdown link checks, npm pack hygiene, and secret-pattern smoke scan.
• opencode install hardened with a subagent frontmatter sanitizer.
• tldr-shrink MCP spawn behavior fixed for Windows .cmd shim resolution.

Install

npx -y github:jqbit/TLDR -- --all

Prompt-only:

curl -fsSL https://raw.githubusercontent.com/jqbit/TLDR/main/install.sh | bash -s --

Verification

Local and GitHub CI passed:

• python3 tests/verify_repo.py
• python3 bench/check-doc-sync.py
• python3 bench/check-md-links.py
• Python safety/hook/validate suites
• node --test tests/installer/*.test.mjs
• standalone JS suites
• npm pack hygiene
• secret-pattern smoke scan

v0.19.0 — canonical TLDR.md only

What's new

• Canonical install now uses the single shipped prompt file: TLDR.md.
• Removed legacy variants: TLDR.blunt.md, TLDR.accurate.md, and TLDR.merged.md.
• Installer now defaults to TLDR.md and keeps optional --with-hermes merge mode.
• Updated docs, issue/PR templates, CI checks, and agent-path mappings for the single-prompt flow.

For full diff, see commit: 4320621.

v0.14.2 — 671 chars, surgical add-backs

Final surgical add-backs over v0.14.1 (tier2): 'no fence wrap', '(when explicitly asked)', '(exact match only)' — recovers edge-case nuance with minimal char penalty. Result: 671 chars (-56% vs v0.13.1).

v0.14.1 — tier2 TAUT (765 chars)

Swap to tier2 TAUT — aggressive prose compression, -50% vs v0.13.1.

v0.14.0 — 11-harness sweep, repo restructure

Major release: TAUT.md surgically compressed to 1290 chars, explicit communication-only scope, output-only override added. Repo restructured to TAUT.md + README.md + LICENSE + data/ + bench/. SVG visualizations + reproducible bench tooling included.

TAUT v0.13 — first public release

Headline

A communication-style system prompt for coding agents that compresses prose output by 80% on average across 8 different agent CLIs while keeping a professional senior-engineer register.

Agent	Δ % vs baseline	Compliance
gemini	−96.0 %	100 %
droid	−85.8 %	100 %
pi	−79.0 %	100 %
openclaw	−78.6 %	100 %
codex	−77.4 %	93 %
claude	−76.2 %	100 %
hermes	−75.2 %	87 % *
cursor-agent	−71.3 %	93 %
TOTAL	−80.0 %	avg 96.7 %

* hermes ceiling at 87% is harness-bounded — its CLI binary forcibly emits diff views and session_id: trailers no prompt can suppress. See EVOLUTION.md §8.

Beats caveman

Caveman's published average is ~65% reduction. TAUT lands at 80%. Beat by 15 percentage points while keeping a professional senior-engineer register (no persona collapse, no character roleplay, no "ooga booga" responses).

What's inside

• TAUT.md — the prompt. Drop into your agent's global instruction file.
• AGENT-LOCATIONS.md — per-agent deploy paths + one-shot bash script.
• README.md — entry point + quick install.
• PHILOSOPHY.md — design rationale, ML grounding (Constitutional AI, InstructGPT, Sycophancy paper, Role-Play LLMs in Nature, etc.), full citations.
• EVOLUTION.md — version-by-version journey from v0.1 (−34 %, 27 % compliance floor) to v0.13 (−80 %, 87 % compliance floor). Variance-shrinkage centerpiece (66 pp → 13 pp = 5× reduction in cross-agent variance).
• BENCHMARKS.md — full per-version per-agent raw data for plotting / analysis.

Quick install

```bash

Drop TAUT.md into each agent's global instruction file

cp TAUT.md ~/.claude/CLAUDE.md
cp TAUT.md ~/.codex/AGENTS.md
cp TAUT.md ~/.gemini/GEMINI.md
cp TAUT.md ~/.factory/AGENTS.md
cp TAUT.md ~/.pi/agent/AGENTS.md
cp TAUT.md ~/AGENTS.md

For openclaw + hermes: append after their existing system prompt — see AGENT-LOCATIONS.md

```

Methodology

• 8 production coding-agent CLIs (Claude Code, Codex, Gemini, Droid, Cursor, Pi, Hermes, OpenClaw)
• 15-prompt frozen suite covering 13 distinct verbosity-trap categories
• N=3 trials per (agent, prompt) for trial-stable agents
• ~3 900 measured agent responses across the v0.1 → v0.13 iteration
• Cross-agent fair tokenizer: `tiktoken o200k_base`
• Strict ALL-trials-pass compliance metric (variance-aware)

Credit

Built on the work of caveman by Julius Brussee. TAUT diverges by replacing the caveman persona with a senior-engineer register — see PHILOSOPHY.md §3 for the full design-divergence rationale.