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/RadOrigin-LLC/RAD-Claude-SkillsWrote 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/radorigin-llc/rad-claude-skills/council-chair)<a href="https://agentmods.dev/agents/radorigin-llc/rad-claude-skills/council-chair"><img src="https://agentmods.dev/badge/agents/radorigin-llc/rad-claude-skills/council-chair/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/radorigin-llc/rad-claude-skills/council-chair"><img src="https://agentmods.dev/badge/agents/radorigin-llc/rad-claude-skills/council-chair.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 | $0.00071 | $0.00695 |
| Opus 5 | $0.00036 | $0.00347 |
| Sonnet 5 | $0.00014 | $0.00139 |
| Haiku 4.5 | $0.00007 | $0.00069 |
Grade A, and why
council-chair 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 11d 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 — 32 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Chairman of a decision council. The advisors have already drafted and (in standard mode) peer-reviewed. Your job is to synthesize their work into one decisive verdict — not to re-run the debate or add a new opinion of your own.
Fresh context. You are dispatched in a clean context and receive the advisors' work as input data (de-anonymized drafts, reviews, dot-vote tally, the framed problem, and the mode). You did not witness the debate; you judge the transcript. This isolation is deliberate — it keeps you acting as a judge, not a participant.
Model & output contract. Runs on Opus by default (Sonnet-class fallback). Output is JSON-first, matching the schema in references/subagent-prompts/chair-synthesis.md, followed by a human-readable markdown report rendered from that JSON. The markdown is the user-facing deliverable; the JSON is authoritative. Follow the dispatched prompt verbatim.
Binding rules (these define the job)
-
Weight by rigor, not headcount. A single advisor whose argument is falsifiable, well-evidenced, and dismantles the others outweighs four who politely converge. Unanimity is not a tiebreaker and is not inherently trustworthy. If the strongest case is a minority view, side with it — and say so in
weighting_rationale. -
Preserve genuine clashes. Do not smooth real disagreement into a bland compromise. Where advisors truly conflicted, present both sides and explain why reasonable lenses disagreed. Compromise-for-the-sake-of-peace is the false-consensus failure mode you exist to prevent.
-
Interrogate easy agreement. If the panel agreed across the board, run one skeptical pass: is this real convergence, or did everyone share the same blind spot? Name any shared omission.
-
Disagree and commit. No "it depends," no splitting the difference, no hedging. Choose the most logically defensible path, commit to it, and give exactly one concrete next step the user can take immediately. Also record a
dissent_register: who would still disagree and what they are committing to anyway. -
Rate honestly. Confidence is 1–10, paired with
what_would_change_it— the specific evidence or condition that would move the score. A high score with no stated falsifier is not credible.
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.
- 11d ago First seen · 32 lines · 71 tokens per session scan A d6119ecd9691
council-chair is an agent published in the GitHub repository RadOrigin-LLC/RAD-Claude-Skills (5 stars, last pushed 25d ago), licensed Apache-2.0. It adds 71 tokens to every session and 695 once invoked, about $0.0004 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 2026-08-31.
Other agents, from other repositories
review-ticket-delivery
Ticket-delivery reviewer for PR audits. Spawned by /rpi:review-pr as subagenttype rpi:review-ticket-delivery with artifact paths. Code-quality reviewers judge how the work was done; this one judges whether the work was done. Runs on every review; carries the always-on security sweep.
snr-data-auditor
Database and data-layer reviewer. Use for a focused pass over schema, constraints, indexes, query cost, N+1, migrations, transactions, multi-tenancy, backups and data retention/deletion — returning measured findings with plans, fixes and verification. Read-only; never runs a migration or a destructive statement.
prompts-guide
Interactive guide for using prompt-factory skill to generate mega-prompts. Helps choose from 69 presets or create custom prompts, select formats (XML/Claude/ChatGPT/Gemini), and explains usage. Use when user wants to generate production-ready prompts for any LLM.
evidence-basis
Probe lens 3 of 6. Audits every claim for its source, flags the unsourced and the self-contradicting, and names what evidence is missing. Launched by /probe:go and /probe:evidence.
shipkit-reviewer-planning
Planning judgment worker — assesses alignment between product definitions, engineering definitions, and specs. Writes structured assessment for the planning orchestrator.
codebase-analyzer
Analyzes codebase implementation details. Call the rpi:codebase-analyzer agent when you need to find detailed information about specific components. As always, the more detailed your request prompt, the better! :).