{
  "markdown": "<p align=\"center\">\n  <img src=\"assets/logo.svg\" alt=\"pager logo\" width=\"800\">\n</p>\n\n# pager\n\n> Ultra-short communication mode for AI coding agents: words &lt;= 5 characters, zero fluff, 100% technical intent preserved.\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)\n[![Compatibility](https://img.shields.io/badge/harnesses-Claude%20%7C%20Gemini%20%7C%20Antigravity%20%7C%20Codex%20%7C%20Cursor-emerald)](#installation)\n[![Token Cut](https://img.shields.io/badge/token_cut-~66%25-brightgreen)](#benchmarks--evals)\n\n`pager` reduces token usage and scanning latency by enforcing a 5-character word limit on conversational output while keeping code blocks, paths, commands, and diagnostic data completely intact.\n\n---\n\n## Comparison\n\n| Standard Response | Pager Mode |\n| :--- | :--- |\n| \"The PostgreSQL connection pool has been exhausted due to unclosed database connections in the authentication middleware. Please restart the auth service immediately and review active query pool limits.\" | `SEV1 auth: DB pool full -> 504 err. Plz rest auth ASAP. Chk qry load.` |\n| **34 words / 48 tokens** | **13 words / 16 tokens (~66% reduction)** |\n\nCode blocks, diffs, terminal commands, and identifiers remain unchanged.\n\n---\n\n## Installation\n\n### 1. Universal Skills CLI\n\nInstall for any skills-compatible agent environment:\n\n```bash\nnpx skills add gauravsaini/pager\n```\n\n### 2. Native Harness Support\n\n- **Google Antigravity**:\n  ```bash\n  agy plugin install ./pager\n  ```\n  *(or `npx skills add gauravsaini/pager -g`)*\n- **Claude Code**:\n  ```bash\n  claude plugin marketplace add gauravsaini/pager && claude plugin install pager@pager\n  ```\n- **Gemini CLI**:\n  ```bash\n  gemini extensions install https://github.com/gauravsaini/pager\n  ```\n- **Cursor IDE**:\n  ```bash\n  npx skills add gauravsaini/pager -a cursor -g\n  ```\n  *(or use the included [.cursorrules](.cursorrules))*\n- **Codex CLI**:\n  ```bash\n  npx skills add gauravsaini/pager -a codex\n  ```\n  *(or use the included [AGENTS.md](AGENTS.md))*\n  For Codex's native GitHub skill installer, use:\n  ```bash\n  python3 ~/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py --repo gauravsaini/pager --path skills/pager\n  ```\n- **Windsurf**:\n  Use the included [.windsurfrules](.windsurfrules).\n- **GitHub Copilot**:\n  Use the included [.github/copilot-instructions.md](.github/copilot-instructions.md).\n\nFor complete harness configuration details, see [INSTALL.md](INSTALL.md).\n\n---\n\n## Usage\n\n- **Trigger**: Run `/pager`, invoke `$pager`, or request `\"pager mode\"`.\n- **Revert**: Say `\"full text\"`, `\"normal mode\"`, or `\"verbose\"` to disable.\n- **Codex auto-start**: Add a short note to `~/.codex/AGENTS.md` to use the\n  installed `pager` skill by default in every conversation.\n- **Antigravity auto-start**: Add `~/.gemini/config/rules/pager.md` (or `.agents/rules/pager.md`)\n  with `trigger: always_on` to activate pager mode automatically in every conversation.\n\n---\n\n## Repository Layout\n\n```\npager/\n  assets/logo.svg                  # Vector branding banner\n  plugin.json                      # Antigravity plugin manifest\n  SKILL.md                         # Root shim for repo-root skill installers\n  skills/pager/SKILL.md            # Canonical skill definition\n  rules/                           # Universal and Antigravity rule definitions\n  .claude-plugin/                  # Claude Code plugin manifest & marketplace definition\n  gemini-extension.json            # Gemini CLI extension manifest\n  AGENTS.md                        # Universal agent instructions (Codex, Antigravity, OpenHands)\n  CLAUDE.md                        # Claude Code rules\n  GEMINI.md                        # Gemini CLI rules\n  .cursorrules                     # Cursor rules\n  .windsurfrules                   # Windsurf rules\n  .github/copilot-instructions.md  # GitHub Copilot custom instructions\n  INSTALL.md                       # Comprehensive installation guide\n  evals/test_pager.py              # Word-length & intent test suite\n  benchmarks/run.py                # English benchmark suite (10 tasks)\n  benchmarks/run_zh.py             # Chinese benchmark suite (5 tasks)\n  benchmarks/prompts.json          # Shared benchmark prompts\n```\n\n---\n\n## Benchmarks & Evals\n\nRun the evaluation and benchmark suites using `uv`:\n\n```bash\n# Run unit tests and intent validation\nuv run python evals/test_pager.py\n\n# Run English benchmark suite\nuv run python benchmarks/run.py\n\n# Run Chinese benchmark suite\nuv run python benchmarks/run_zh.py\n\n# Dry run (list tasks without executing)\nuv run python benchmarks/run.py --dry-run\n```\n\n- **English Benchmark**: ~66% token reduction compared to baseline, 10/10 compliance.\n- **Chinese Benchmark**: ~51% character reduction, 5/5 pass.\n\n---\n\n## License\n\nMIT. See [LICENSE](LICENSE).\n",
  "bytes": 4813,
  "sha": "de40de249bdaddeb8b5e1218d0fa8fd4b10942ca688c6726dab15ed9cece6bb4",
  "repo_slug": "gauravsaini/pager",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_gauravsaini_pager_3d1b087a/readme"
}