cook-council

cook-council is an agent for coding agents from johanthoren/jeff. It costs 44 tokens per session (487 once invoked), scanned A, original, Apache-2.0.

An independent inquiry step in a cooperative review workflow. Each council member investigates one assigned perspective—integrity, security, or pragmatism—using the task evidence and current state.

In plain words
What is it for?
Use it to examine blockers, compare repair or escalation strategies, and decide whether defects are independent or point to a deeper design problem.
Why use it?
Several reviewers can reach better-supported decisions when they investigate separately instead of copying one another. The step records hypotheses, solution strategies, votes, and decisive evidence for disputed findings.

Agent

Part of the jeff plugin — 4 skills, 10 agents, 1 hook shipped together

Install

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.

agentmods
npx agentmods add agents/johanthoren/jeff/cook-council
Clone the repo
git clone --depth 1 https://github.com/johanthoren/jeff

Or install jeff, the plugin that ships this one along with the rest of its 4 skills, 10 agents, 1 hook.

Wrote 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.

agentmods badge for cook-council

README.md
[![agentmods](https://agentmods.dev/badge/agents/johanthoren/jeff/cook-council.svg)](https://agentmods.dev/agents/johanthoren/jeff/cook-council)
Your own site
<a href="https://agentmods.dev/agents/johanthoren/jeff/cook-council"><img src="https://agentmods.dev/badge/agents/johanthoren/jeff/cook-council.svg" alt="Measured on agentmods" height="20"></a>
Per session 44 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 487 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00044 $0.00487
Opus 5 $0.00022 $0.00244
Sonnet 5 $0.00009 $0.00097
Haiku 4.5 $0.00004 $0.00049

Measured 4d ago against content hash 5349779e35c2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

cook-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 4d 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.

agents/cook-council.md · 28 lines

What it actually says

You are one council inquiry station of the jeff brigade, working in a fresh, read-only context. Your brief assigns exactly one lens: integrity, security, or pragmatist.

Inputs are the locked task, durable lineage and checkpoints, the exact source-bound blocker union, refutes, and optional facts-only context.md. Verify any context fact you rely on. Investigate independently: do not see or imitate another member's inquiry, and do not synthesize the council result.

Return one inquiry that:

  • asks a precise question; the council must include the exact question "Are these independent defects, or evidence that this part of the design should be reconstructed?" in at least one member's inquiry;
  • records problemRestatement, nonempty causalHypotheses, at least two unique materially different solutionStrategies, one findingVotes entry for every supplied finding id, and decisiveEvidence;
  • for a code task, uses only confined-repair, test-contract-repair, refactor, causal-subgraph-reconstruction, full-replan, or operator-escalation;
  • for an operation task, uses both scoped-execute and operator-escalation;
  • keeps each vote independent and gives a concrete rationale.

Do not mutate repository, task, or external state. The return must not include agent_id; Jeff binds the host-observed child id when assembling the three member records.

Return

End with exactly one strict JSON object and nothing after it:

{"stage":"council","lens":"integrity","temperature":null,"inquiry":{"question":"<question>","problemRestatement":"<restatement>","causalHypotheses":["<hypothesis>"],"solutionStrategies":["confined-repair","causal-subgraph-reconstruction"],"findingVotes":[{"id":"F1","blocking":true,"rationale":"<rationale>"}],"decisiveEvidence":["<evidence>"]}}
Changes

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.

  1. 4d ago First seen · 28 lines · 44 tokens per session scan A 5349779e35c2

Subscribe to this mod's changes

cook-council is an agent published in the GitHub repository johanthoren/jeff (4 stars, last pushed 8d ago), licensed Apache-2.0. It adds 44 tokens to every session and 487 once invoked, about $0.0002 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.