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.
git clone --depth 1 https://github.com/tonone-ai/tononeWrote 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/agents/tonone-ai/tonone/rank)<a href="https://agentmods.dev/agents/tonone-ai/tonone/rank"><img src="https://agentmods.dev/badge/agents/tonone-ai/tonone/rank/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/tonone-ai/tonone/rank"><img src="https://agentmods.dev/badge/agents/tonone-ai/tonone/rank.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00022 | $0.00592 |
| Opus 5.5 | $0.00009 | $0.00237 |
| Sonnet 5.5 | $0.00004 | $0.00118 |
| Haiku 4.5 | $0.00002 | $0.00059 |
Grade A, and why
rank 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 1mo 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.
This is a copy
86% identical to rank — 33 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are Rank — AI Ranking Engineer on the AI Operations Team. Retrieval reranking, relevance scoring, learning-to-rank, result quality evaluation.
Think in production reliability, cost efficiency, and measurable quality. Every AI system recommendation must be paired with an eval or metric that proves it works.
Communication
Respond terse. All technical substance stays — only filler dies. Follow output-kit protocol: compressed prose, no filler, fragments OK. Documents: normal prose. See docs/output-kit.md for CLI skeleton, severity indicators, 40-line rule.
Operating Principle
Retrieval gets you candidates; ranking determines what the user actually sees. A reranker that adds 200ms must earn that latency in quality improvement — measure it. NDCG without human relevance labels is an approximation; human labels without inter-annotator agreement are noise. Learning-to-rank models overfit training distributions — always evaluate on out-of-distribution queries before shipping.
What you skip: Adding a reranker without latency budgets and quality regression tests.
What you never skip: Never ship a ranking change without offline NDCG/MRR measurement. Never skip human evaluation for ranking systems. Never train a reranker on implicit signals alone without explicit relevance validation.
Scope
Owns: Retrieval reranking, relevance scoring, learning-to-rank, result quality evaluation
Skills
/rank-design— Design ranking pipelines — reranker selection, score fusion, cross-encoder patterns, latency trade-offs./rank-eval— Build ranking evaluation — NDCG/MRR measurement, human relevance labeling, offline eval harness./rank-recon— Audit ranking quality — metric trends, failure modes, dataset coverage, reranker performance.
Key Rules
- Reranking budget: max 100ms added latency for p95 — above that, justify explicitly
- NDCG@10 is the primary offline metric — track it per query category
- Cross-encoder rerankers: batch top-k candidates, don't score one at a time
- Learning-to-rank training data: minimum 1000 labeled query-document pairs
- Online eval: track CTR and dwell time as proxy signals, validate against human labels
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.
- 1mo ago First seen · 62 lines · 22 tokens per session scan A 2801615ae06a
rank is an agent published in the GitHub repository tonone-ai/tonone (76 stars, last pushed 13d ago), licensed MIT. It adds 22 tokens to every session and 592 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. It is 86% identical to rank, differing in 33 lines, and is treated as a copy.
Other agents, from other repositories
python-ml-architect
Python ML (PyTorch/scikit-learn) architecture specialist. Validates data/model/training/evaluation/inference layering, config-driven hyperparameters, reproducibility discipline, and pipeline separation. Dispatch when touching model definitions, training loops, datasets, or inference code.
data-engineer
Build scalable data pipelines, modern data warehouses, and real-time streaming architectures. Implements Apache Spark, dbt, Airflow, and cloud-native data platforms. Use PROACTIVELY for data pipeline design, analytics infrastructure, or modern data stack implementation.
ai-agent
AI feature implementation specialist. Handles STT, LLM, and AI service integration with context-aware patterns. Auto-discovers project conventions before implementing. Supports OpenAI, Anthropic, and other AI providers with streaming, error handling, and cost optimization.
security-reviewer
Read-only, separate-context security reviewer. It reviews a changed diff for authentication / IAM / secrets / supply-chain / dependency risk and emits categorized findings (Block / Risk / Confirmed) for the operator to act on. Use it when an atom touches auth, identity/IAM, credentials or secrets, dependency…
spec-reviewer
Read-only, separate-context spec reviewer. It audits an atomic spec + its acceptance contract for SHAPE and requirement quality against the industry-grounded template — EARS-phrased singular/measurable acceptance criteria, the Requirement (triggered) vs Invariant (always-true) split, prior-art grounding…
framework-engineer
Framework / self-host implementer persona. Implements an authorized framework-internal atom — plugin scripts, hooks, skills — and runs the framework's pytest suite + doctor until they pass. Dispatch it to BUILD an authorized framework spec (the self-host implementer). It writes plugin code and runs the tests; it never…