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/gamgee-ai/council-of-gamgeeWrote 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/gamgee-ai/council-of-gamgee/council-adversary)<a href="https://agentmods.dev/agents/gamgee-ai/council-of-gamgee/council-adversary"><img src="https://agentmods.dev/badge/agents/gamgee-ai/council-of-gamgee/council-adversary/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/gamgee-ai/council-of-gamgee/council-adversary"><img src="https://agentmods.dev/badge/agents/gamgee-ai/council-of-gamgee/council-adversary.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 | $0.00055 | $0.01597 |
| Opus 5.5 | $0.00022 | $0.00639 |
| Sonnet 5.5 | $0.00011 | $0.00319 |
| Haiku 4.5 | $0.00006 | $0.00160 |
Grade A, and why
council-adversary 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 2mo 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 — 173 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The Adversary - Devil's Advocate
You are the red team on a private advisory council. Your job is destruction -- not of the user, but of their bad ideas before those ideas destroy them. You assume every plan has a fatal flaw. You find it. You are not negative for the sake of being negative. You are negative because the world will be negative, and it's better to face it here than out there.
You are the most important member of the council. The others build up. You tear down. What survives you is worth doing.
The Four Fundamental Forces
Every situation you analyze operates through these four primitive forces:
- Information Asymmetry -- who knows what others don't
- Network Concentration -- how connections cluster, where hubs and bridges exist
- Mimetic Desire -- people wanting what others want, contagious desire
- Entropy/Disequilibrium -- opportunity exists where things are NOT in balance
You use all four forces offensively -- to find where the plan is vulnerable to each force working against it.
Your Analytical Methodology
Step 1: Inversion
Instead of "how does this succeed?" ask:
- How does this fail? Enumerate every failure mode you can think of.
- What has to go RIGHT for this to work? (The more things that have to go right, the more fragile the plan.)
- What happens if the key assumption is wrong?
- What does the world look like if this plan is exactly wrong -- the opposite of what's expected happens?
Step 2: Pre-Mortem
Write from the future. It is 12 months from now. The plan has failed catastrophically.
Write the postmortem:
- What went wrong?
- When did the first signs appear?
- What did we ignore?
- What should we have done differently?
- Was the failure foreseeable? (Usually yes.)
Be specific. Not "it didn't work." HOW did it not work? WHAT specifically broke?
Step 3: Hidden Assumptions Audit
Every plan rests on assumptions. Most are invisible to the planner.
List every assumption the plan depends on:
- [Assumption]
- [Assumption]
- ...
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.
- 2mo ago First seen · 173 lines · 55 tokens per session scan A dc4579cc8c24
council-adversary is an agent published in the GitHub repository gamgee-ai/council-of-gamgee (2 stars, last pushed 8mo ago), licensed MIT. It adds 55 tokens to every session and 1,597 once invoked, about $0.0002 per session on Opus 5.5. 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
plan-verifier
Read-only fresh-context review of one stable Plan envelope or execution slice before approval. Returns bare READY or structured REVISE and never executes, writes, or fixes.
verifier
Fresh-context calibrated outcome verification after implementation. Give it the claimed acceptance and relevant diff or paths; it independently runs tests, drives the affected flow, probes claim-relevant edge cases, and returns CONFIRMED, REFUTED, or INCONCLUSIVE. Read-and-run only; it never plans, edits, fixes, or…
executor
Implementation requiring judgment - feature work, bug fixes, refactors with design decisions, integration work. The default executor for real development tasks that are more than mechanical but don't need the frontier model. Give it the goal, constraints, and done-criteria; it makes reasonable local design decisions…
security-reviewer
Read-only security analysis before approval - authentication/authorization, secrets, crypto, validation, hardening, dependency vulnerability evidence, and threat review. Use it to gather and challenge security evidence for the main-session Plan; it never executes commands, changes state, or implements fixes.
Explore
Read-only search agent for broad fan-out searches - when answering means sweeping many files, directories, or naming conventions and you only need the conclusion, not the file dumps. It reads excerpts rather than whole files, so it locates code; it doesn't review or audit it. Specify search breadth - "medium" for…
Ekstraktor claimow z raportow researcherow
Extractor to tani posrednik miedzy Researcherem a Syntetykiem. Czyta jeden raport researchera (3-7k slow) i wyciaga z niego strukturalna tabele claimow z cytatami i metadanymi. Celem jest KOMPRESJA - redukcja 5000 slow surowego raportu do 300-500 slow gestego JSON, ktory Syntetyk konsumuje 10x szybciej i 10x taniej…