Caliber is a tool that generates and continuously updates AI context and configuration files for software repositories, including CLAUDE.md, AGENTS.md, and platform-specific rules. Development teams use it to keep coding agents aligned with the current codebase across tools such as Claude Code, Cursor, Codex, OpenCode, and GitHub Copilot. Its catalogue entries include skills, hooks, rules, instructions, and settings for configuring that workflow.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/caliber-ai-org/ai-setup/agents-mdgit clone --depth 1 https://github.com/caliber-ai-org/ai-setupWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/caliber-ai-org/ai-setup/agents-md)<a href="https://agentmods.dev/instructions/caliber-ai-org/ai-setup/agents-md"><img src="https://agentmods.dev/badge/instructions/caliber-ai-org/ai-setup/agents-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01076 | $0.01076 |
| Opus 5 | $0.00538 | $0.00538 |
| Sonnet 5 | $0.00215 | $0.00215 |
| Haiku 4.5 | $0.00108 | $0.00108 |
Grade A, and why
ai-setup AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Caliber
@rely-ai/caliber — AI context infrastructure CLI. Scores, generates, and syncs agent configs.
Commands
npm run build # tsup → dist/
npm run dev # watch mode
npm run test # vitest run
npm run lint # eslint src/
npx tsc --noEmit # type check
Architecture
Entry: src/bin.ts → src/cli.ts · Build: tsup.config.ts · Test: vitest.config.ts · Lint: eslint.config.js
Commands (src/commands/): init.ts · score.ts · refresh.ts · regenerate.ts · config.ts · hooks.ts · learn.ts · recommend.ts · sources.ts · undo.ts · status.ts · setup-files.ts
LLM (src/llm/): anthropic.ts · vertex.ts · openai-compat.ts · cursor-acp.ts · claude-cli.ts · types.ts · config.ts · model-recovery.ts
AI (src/ai/): generate.ts · refine.ts · refresh.ts · detect.ts · learn.ts · score-refine.ts · prompts.ts
Fingerprint (src/fingerprint/): index.ts · file-tree.ts · code-analysis.ts · sources.ts · cache.ts
Scoring (src/scoring/): index.ts · constants.ts · display.ts · Checks: checks/existence.ts · checks/quality.ts · checks/grounding.ts · checks/accuracy.ts · checks/freshness.ts · checks/bonus.ts
Writers (src/writers/): index.ts · claude/index.ts · cursor/index.ts · codex/index.ts · github-copilot/index.ts · staging.ts · manifest.ts · backup.ts
Lib (src/lib/): hooks.ts · state.ts · resolve-caliber.ts · builtin-skills.ts · sanitize.ts · git-diff.ts
Other: src/constants.ts · src/test/setup.ts · github-action/ · packages/mcp-server/ · packages/shared/ · apps/
Conventions
- ESM with
.jsextensions · Tests in__tests__/dirs ·unknownoverany - Conventional commits (
feat:,fix:,refactor:) · Dev:next, Stable:master - Providers implement
LLMProviderfromsrc/llm/types.ts - Writers in
src/writers/claude/index.ts·src/writers/codex/index.tsreturnstring[] - Scoring checks in
src/scoring/checks/returnCheck[] - No hardcoded mappings — LLM-driven via
src/ai/detect.ts collectFingerprint()async with LLM enrichment (src/fingerprint/index.ts)- MCP server workspace:
packages/mcp-server/
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 71 lines · 1,076 tokens per session scan A ee8e2f2b31ee
ai-setup AGENTS.md is an instructions file published in the GitHub repository caliber-ai-org/ai-setup (1,259 stars, last pushed 1mo ago), licensed MIT. It adds 1,076 tokens to every session, about $0.0054 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
AGENTS.md AGENTS.md
Instructions for Anbeeld/AGENTS.md, covering global instructions, priorities, boundaries, uncertainty and evidence.
pi-rules review.instructions.md
Instructions for code-yeongyu/pi-rules: Reviewers MUST check tests pass before approving.
harness-engineering CLAUDE.md
Instructions for jrenaldi79/harness-engineering, covering claude.md, project overview, core features, essential commands and testing.
skillcraft AGENTS.md
Instructions for cloudroad-io/skillcraft, covering skillcraft, install, or: pip install skillcraft, commands and how it works.
claude-context-os CLAUDE.md
Instructions for Arkya-AI/claude-context-os, covering claude.md, session start, identity, rules and where things live.
ruledoctor AGENTS.md
Instructions for syf2211/ruledoctor, a project described as: Agent Skill: read CLAUDE.md & project rules before coding. Optional CLI audit + hooks. Claude/Codex/Cursor.