llms.txt@site/public git:20261016.8aedb22A

Immutable. This exact content never changes and is served at /api/v1/blob/fa76d3392c7242d7.

# Chisle

> Chisle is a token-efficiency plugin for AI coding agents that cuts LLM token usage on three axes: (1) a terse senior-dev persona that writes maximum-signal prose and YAGNI-first code, (2) a post-tool hook that compresses tool output (ANSI scrub, head+tail elide with error-line salvage, same-session dedup) before it enters the context window, and (3) context-diet rules that teach the model to read less (grep before Read, slice reads, filter at source). Claude Code and Pi run all three axes plus live /chisle toggling and a savings badge; Cursor, Windsurf, Cline, Kiro, Antigravity, Codex, Gemini, and GitHub Copilot receive the ruleset via generated rule files; OpenCode gets the global fenced ruleset plus on-demand skills; Hermes Agent gets the skills as /chisle commands in ~/.hermes/skills.

Key measured results (same model and prompts, only the injected ruleset differs; reproducible via benchmarks/ in the repo):
- Claude Code (Haiku 4.5, 26 live prompts, tools off): answer length Chisle 56% of a bare model, ponytail 58%, caveman 65%. Answer lines Chisle 52%, ponytail 54%, caveman 66%. Prompts that got longer: Chisle 12/26, caveman 13, ponytail 18. Worst prompt: Chisle 194%, caveman 339%, ponytail 493%.
- Pi + GPT-5.5 (6 tasks): answer length Chisle 37%, ponytail 49%, caveman 56%. Billed tokens on explanation prompts Chisle 39%, caveman 43%, ponytail 76%.
- Tool output is 67.5% of session content across 171 real Claude Code transcripts; the compressor trims eligible outputs ~46%. On OpenCode a 30 KB log cut fresh input tokens 87%.
- Quality: 156/156 on auto-graded maths and coding problems (bare model 155/156).
- Plugin overhead: ~900 tokens of rules, served from the prompt cache after the first request, plus a ~50-token per-turn reminder.

Safety/correctness: the compressor runs behind an allowlist (Claude Code: Bash, Agent, WebFetch, WebSearch, Grep, Glob, mcp__*; Pi: bash, powershell, grep, find, ls, mcp__*; Copilot CLI: bash, powershell, grep, glob, web_fetch, web_search, task; MCP proxy: any tool not named like read/view/file/edit/write/patch) and never modifies Read/Edit/Write results, whose exact bytes feed later edits. Deterministic, zero dependencies, zero network calls, zero LLM calls. One mode, no intensity levels. Kill switches: CHISLE_COMPRESS=0, "stop chisle", /chisle off, npx chisle --uninstall.

Install: `npx chisle` (npm, auto-detects all 13 agents; `npx chisle mcp wrap` adds MCP output compression to Cursor, Windsurf, Cline, Kiro, Gemini, Qwen Code and Copilot CLI), `claude plugin marketplace add JayPokale/Chisle` then `claude plugin install chisle@chisle` (Claude Code plugin), or `pi install npm:chisle` (Pi package). Requires Node >= 18. MIT license.

Lineage: descendant of the caveman and ponytail projects (terse-persona axis), both credited in the repository's prior-art table and NOTICE file; Chisle adds the tool-output-compression and context-diet axes and publishes live benchmarks against both.

## Links

- [GitHub repository](https://github.com/JayPokale/Chisle): source, benchmarks, raw transcripts
- [npm package](https://www.npmjs.com/package/chisle): installer
- [Benchmarks](https://github.com/JayPokale/Chisle/tree/main/benchmarks): 4-arm live runner + results
- [Changelog](https://github.com/JayPokale/Chisle/blob/main/CHANGELOG.md)

## Author

Jay Pokale (https://github.com/JayPokale), co-engineered with Claude (Anthropic), Codex (OpenAI), and Antigravity (Google).

Pi (added 3.1.0): Pi exposes a tool_result event, so the input axis runs there too. Pi already truncates built-in tool output at 50KB/2000 lines, so the honest figure is the marginal saving on top of that: 27.4% of persisted tool-output characters across 8,044 results. On the live six-task output-axis run Chisle billed 61% of vanilla and lost billed output to ponytail at 59%, while producing the smallest visible answers (37%). Both numbers are committed in benchmarks/results/2026-09-11-pi.md.