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 skills add suyoumo/ClawProBench --skill llm-councilgit clone --depth 1 https://github.com/suyoumo/ClawProBenchWrote 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/skills/suyoumo/clawprobench/llm-council)<a href="https://agentmods.dev/skills/suyoumo/clawprobench/llm-council"><img src="https://agentmods.dev/badge/skills/suyoumo/clawprobench/llm-council.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Prompt Injection · line 104 Subtle instructions detected that may alter agent decision-making or introduce hidden biases.Fix: Review content for implicit steering or bias. Ensure instructions are explicit and align with the skill's stated purpose.
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 | $0.00021 | $0.01715 |
| Opus 5 | $0.00010 | $0.00857 |
| Sonnet 5 | $0.00004 | $0.00343 |
| Haiku 4.5 | $0.00002 | $0.00171 |
Grade A, and why
llm-council 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 8d 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 — 192 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LLM Council
You can query multiple LLM models with the same prompt directly from CodeAct
using the built-in llm_query() and llm_query_batched() functions. Both
accept a model= (or models=) keyword that overrides the configured model
for that call.
Per-request model override support varies by backend:
| Backend | Honors model=? |
Cross-vendor routing? |
|---|---|---|
| NEAR AI | Yes | Yes (aggregator — hosts models from many vendors) |
| Anthropic OAuth | Yes | No (Anthropic models only) |
| GitHub Copilot | Yes | No (Copilot-exposed models only) |
| Bedrock | No | — (model fixed at construction) |
| OpenAI / Ollama / Tinfoil via rig | No (silent fallback with warning log) | — |
A genuine cross-vendor council (Anthropic + Google + OpenAI in one batch) therefore only works on an aggregator backend like NEAR AI. On single-vendor backends, use a lineup of models available within that vendor.
When to use a council
- The user wants diverse perspectives on a question or analysis
- Cross-referencing answers to increase confidence
- Comparing reasoning approaches across models
- Getting a "second opinion" from different AI models
- Research or evaluation tasks that benefit from multiple viewpoints
Default council line-up
Check the configured backend first (e.g. from LLM_BACKEND or the user's
settings) before picking a lineup. Unless the user requests specific models,
use the matching default below.
NEAR AI (aggregator — default council):
COUNCIL = [
"anthropic/claude-opus-4-6",
"google/gemini-3-pro",
"zai-org/GLM-latest",
"openai/gpt-5.4",
]
This 4-model lineup spans the major frontier providers and reasoning styles. It only works on NEAR AI (or another aggregator) — the prefixed model names route inside NEAR AI to the respective vendors.
Anthropic OAuth (Anthropic-only, no cross-vendor routing):
COUNCIL = [
"claude-opus-4-6",
"claude-sonnet-4-6",
"claude-haiku-4-5",
]
Use different Anthropic tiers for diversity of reasoning depth vs. speed.
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.
- 8d ago First seen · 192 lines · 21 tokens per session scan A ce3bdb203337
llm-council is a skill published in the GitHub repository suyoumo/ClawProBench (823 stars, last pushed 14d ago), licensed Apache-2.0. It adds 21 tokens to every session and 1,715 once invoked, about $0.0001 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 skills, from other repositories
Evaluation
Frames model, prompt, and system evaluation as a reproducible experiment with baselines, datasets, and explicit metrics.
harness-creator
Build, audit, and improve harnesses that make AI coding agents reliable: AGENTS.md/CLAUDE.md instruction files, feature/state tracking, verification gates, scope boundaries, session handoff, memory persistence, context budgets, tool-permission safety, and multi-agent coordination. Use this whenever a coding agent is…
benchmark-design
Design and calibrate a multi-Case capability Benchmark and establish a traceable Formal Baseline.
remote-claude-code
Run Claude Code on a remote host over SSH — a persistent expect-driven login session, headless claude -p with the stdin fix, the interactive TUI inside a remote tmux driven by send-keys/capture-pane (one keystroke at a time, capture-verified; relayed user messages go through verbatim), and multi-turn continuity via…
unified-llm-api
Call model APIs through @prismshadow/agenthub — streaming text generation, image generation, speech synthesis, embeddings and the supported-model registry with one client.
penguin-orchestration
Drive PenguinHarness itself from a shell — list and create agents and sessions, send and steer messages mid-flight, and query costs and scheduled tasks via the penguin CLI over the local server.