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 rules/caliber-ai-org/ai-setup/llm-layergit clone --depth 1 https://github.com/caliber-ai-org/ai-setupWhat 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 | $0.00000 | $0.00195 |
| Opus 5 | $0.00000 | $0.00097 |
| Sonnet 5 | $0.00000 | $0.00039 |
| Haiku 4.5 | $0.00000 | $0.00019 |
Grade A, and why
llm-layer 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 3d 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.
What it actually says
- Providers implement
LLMProviderfromsrc/llm/types.ts:call(),stream(), optionallistModels() - Config: env vars →
~/.caliber/config.jsonviasrc/llm/config.ts - Seat-based:
isSeatBased()insrc/llm/types.ts(cursor, claude-cli) - Cursor:
agent --print --trust --workspace /tmpinsrc/llm/cursor-acp.ts - Fast model:
getFastModel()insrc/llm/config.ts - Model recovery:
src/llm/model-recovery.ts· Errors:src/llm/seat-based-errors.ts - Usage:
trackUsage()fromsrc/llm/usage.ts validateModel()skips seat-based providers
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.
- 3d ago First seen · 13 lines · 0 tokens per session scan A d4ec9a0e8c36
llm-layer is a cursor rule published in the GitHub repository caliber-ai-org/ai-setup (1,259 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 195 tokens. 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 cursor rules, from other repositories
cursorrules
DeepInit projection provenance (R3) stage: EMIT → PROJECT (cursor) runid: run-2026-06-13-kemal-e2e source: AGENTS.md (lean tier) — deterministic projection, no new findings note: content inside the DEEPINIT markers is owned + regenerated; edit OUTSIDE them. -->.
cursorrules
Cursor rule "cursorrules" from syf2211/ruledoctor, covering 订单服务项目规则（demo） and 硬性规则.
integrations
External integration rules — providers, idempotency, bulkhead, observability.
architecture
FastAPI layer architecture rules — Router → Service → Repository with Protocol injection.
main
Core coding standards enforced on all source files.
coding-style
Coding style conventions for C#, TypeScript, Angular, and Python projects.