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/fmflurry/settings-opencodeWrote 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/commands/fmflurry/settings-opencode/arena)<a href="https://agentmods.dev/commands/fmflurry/settings-opencode/arena"><img src="https://agentmods.dev/badge/commands/fmflurry/settings-opencode/arena/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/commands/fmflurry/settings-opencode/arena"><img src="https://agentmods.dev/badge/commands/fmflurry/settings-opencode/arena.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 | $0.00014 | $0.05278 |
| Opus 5.5 | $0.00006 | $0.02111 |
| Sonnet 5.5 | $0.00003 | $0.01056 |
| Haiku 4.5 | $0.00001 | $0.00528 |
Grade A, and why
arena scanned grade A with 1 finding 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 today.
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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s http://localhost:4000/health || echo "Gateway unreachable" How it starts
The opening of the file, as written. The whole thing — 465 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Arena Command
CC: Execute this as a direct CLI wrapper. Run pre-flight checks, execute dry-run first, ask for confirmation, then run live. Read and present results.
Run LLM arena benchmark: $ARGUMENTS
Your Task
Run the LLM arena benchmark to evaluate candidate models on classify/draft prompts. The arena benchmarks each model with multiple criteria (schema_valid, output_quality, latency, etc.) and aggregates results per config.
1. Parse Arguments
From $ARGUMENTS, extract:
prompt:classify|draft|all(default:all)- Optional flags:
--dry-run(automatic in step 2, you run this always first)--confirm(automatic in step 4, only after user approval)--failures(include failures in detail JSON)--max-cost <eur>(budget cap in EUR; exit non-zero if exceeded)--reps <n>(repetitions per case, default: as configured)--top <n>(Stage 1 only: keep top N configs by ranking; default 3)--refine+--from <rankingFile>(Stage 2: dense settings sweep on top-N winners; requires prior ranking.json)--rerank+--from <detailFile>(Offline re-ranking: reload stored detail.json results, re-apply current thresholds (env vars), zero LLM calls, outputs updated ranking.json with provenance via rerankOf metadata)--suggest+--eu(Shortlist generator: no LLM calls, writes candidates proposal; --eu uses eu.openrouter.ai host)
2. Pre-Flight Checks
Stop and report failure if ANY check fails.
Environment Variables:
LITELLM_MASTER_KEY # LiteLLM gateway authentication
OPENROUTER_API_KEY # OpenRouter fallback models
AZURE_AI_API_KEY # Azure models
AZURE_AI_API_BASE # Azure endpoint
ARENA_MAX_COST_EUR # Budget cap in EUR (default: 10)
ARENA_USD_TO_EUR_RATE # Static conversion rate (default: 0.86)
BENCH_FORCE_FALLBACK_CASES # Force rich 25-case fallback datasets (default: 0; set to 1 for baseline runs)
ARENA_CALL_TIMEOUT_MS # Per-call LLM abort timeout in ms (default: 120000; clamp: 10s–5min)
ARENA_TIME_BUDGET_MIN # Global wall-clock budget in minutes (default: 60; hard stop if exceeded)
ARENA_CONCURRENCY # Max concurrent API calls (default: 6; clamp: 1–16)
ARENA_CONCURRENCY_PER_MODEL # Concurrency per model; AIMD throttle on 429 (default: 3; clamp: 1–8)
ARENA_GATE_SCHEMA_VALID_MIN # Schema validation pass-rate minimum (default: 0.98; configurable 0.0–1.0)
ARENA_GATE_VERBATIM_MIN # Verbatim exactness global minimum (default: 0.95; applies to draft montant/numFacture/echeance unless overridden)
ARENA_GATE_VERBATIM_AMOUNT_MIN # Override for montant field (default: ARENA_GATE_VERBATIM_MIN)
ARENA_GATE_VERBATIM_INVOICE_NUMBER_MIN # Override for numeroFacture field (default: ARENA_GATE_VERBATIM_MIN)
ARENA_GATE_VERBATIM_DUE_DATE_MIN # Override for echeance field (default: ARENA_GATE_VERBATIM_MIN)
ARENA_GATE_INTENT_MATCH_MIN # Classify quality gate threshold (default: 0.8; configurable 0.0–1.0)
ARTIFICIALANALYSIS_API_KEY # Optional: enriches --suggest candidate priors
LITELLM_BASE_URL # LiteLLM proxy base URL (default: http://localhost:4000/v1; must use port 4000 with docker-compose.override.yml)
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.
- today Changed a1c4ddd92f9c
- 26d ago First seen · 465 lines · 14 tokens per session scan A 9fa430c2294e
arena is a command published in the GitHub repository fmflurry/settings-opencode (171 stars, last pushed yesterday), licensed MIT. It adds 14 tokens to every session and 5,278 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-20.
Other commands, from other repositories
verify-math
Verify a self-authored mathematical result end to end by routing claims across adversarial review, numerical falsification, symbolic or CAS checks, and Lean, then aggregating one report. Use when a theorem, proposition, conjecture, or paper-wide mathematical argument needs the appropriate combination of verification…
diff
Quantitative volume comparison between a CadQuery model and a reference STEP file.
simulation-calibrator
Test and refine simulation accuracy with validation loops, bias detection, and continuous improvement frameworks.
arg-diagram
ARG academic-paper diagram mode — standalone structural & conceptual diagram generation.
plot
Publication Polish. Runs after results-analysis (Phase 7). Audits all tables and figures produced in Phases 4–7, upgrades them to top-journal standards by calling the table and figure skills.
graphite-morphology-classify
Classify graphite in a cast-iron micrograph per ASTM A247 / ISO 945-1, quantify nodularity, and read the matrix — the single most diagnostic observation in a cast-iron case.