TLDR.md — Too Long Didn't Read

Tiny prompt. Shorter answers. Same brain.

TLDR.md makes AI assistants answer directly: less filler, less fake enthusiasm, less "let me know if..." sludge.

It changes communication style only. Tools, reasoning, code quality, and safety stay the same.

Pick one

File	Use this if...
TLDR.md	You want terse output. Start here.
TLDR.blunt.md	You want terse output plus less sycophancy / more pushback when warranted.
TLDR.accurate.md	You want accurate, complete answers without extreme brevity sacrificing precision.

One-line install

No clone. No editing. The install script writes the chosen prompt to the 7 standard coding-agent locations.

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

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

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

Optional: include Hermes too.

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

--with-hermes preserves an existing ~/.hermes/SOUL.md, makes a backup, and appends or updates a managed TLDR block. Use --overwrite-hermes only if you want prompt-only SOUL.md.

Prefer to inspect the commands instead of piping to bash? Use the manual copy/paste setup in data/agent-locations.md.

For chat apps or web UIs, paste the file into custom instructions, project instructions, system prompt, or a saved prompt.

Verify

The install script prints this automatically, but here is the manual check:

for p in ~/.claude/CLAUDE.md ~/.gemini/GEMINI.md ~/.codex/AGENTS.md \
         ~/AGENTS.md ~/.config/opencode/AGENTS.md \
         ~/.factory/AGENTS.md ~/.pi/agent/AGENTS.md; do
  [ -f "$p" ] && grep -q "^# TLDR" "$p" && echo "✓ $p" || echo "✗ $p"
done

grep -q "target 3 words" ~/.hermes/SOUL.md 2>/dev/null && echo "✓ ~/.hermes/SOUL.md" || echo "✗ ~/.hermes/SOUL.md"

Current defaults

• default: 1 sentence
• target: 3 words
• 1 word when sufficient
• default max: 6 words
• longer only if asked
• greet: 1 word

Example outputs

Port busy; free it.

Yes. Start SQLite.

git reset --soft HEAD~1

What it fixes

• question restatement
• fake enthusiasm / validation
• command wrappers when you asked for only the command
• extra caveats and summary paragraphs
• "let me know if you want more" endings

Benchmarks (historical)

Current prompt sizes:

File	Bytes
TLDR.md	1,165
TLDR.blunt.md	1,487
TLDR.accurate.md	1,627

These benchmark results were measured on earlier shipped prompts. The current prompt files were later tightened to a 1-sentence / 3-word-default / 6-word-max profile and have not yet been rerun through the full bench.

• TLDR.md v0.13.1: −82.1% total prose reduction, 100% average compliance (5 agents × 5 prompts).
• TLDR.md v0.14.3: −80.0% single-turn prose reduction; −75.1% across 8-turn coding conversations; no significant decay.
• TLDR.blunt.md v0.18.0: DSPy round-2 + 5-agent cross-model validation; avg pushback 0.848, correct-user agreement 0.912, mean prose 11.0 words, validation phrases 0%.
• TLDR.accurate.md v0.1.0: Accuracy-first variant; prioritizes correctness and detail where brevity would harm precision.

Full historical details:

• data/agent-locations.md
• data/benchmarks.md
• data/dspy-cross-model-results.md
• data/changelog.md
• CONTRIBUTING.md

License

MIT. See LICENSE.