adversarial-challenger

adversarial-challenger is an agent for coding agents from SSBun/csl-agent-kit. It costs 68 tokens per session (527 once invoked), scanned A, original, MIT.

An independent checking role in a structured answer-review process that looks for errors, missing information, opposing views, risks, trade-offs, and inconsistencies.

In plain words
What is it for?
Use it to challenge a draft across repeated review passes and report stable issue IDs for the coordinator to resolve.
Why use it?
It helps catch problems that the agent writing the answer may overlook.

Agent

Part of the csl plugin — 34 skills, 2 commands, 5 agents, 10 hooks 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/ssbun/csl-agent-kit/adversarial-challenger
Clone the repo
git clone --depth 1 https://github.com/SSBun/csl-agent-kit

Or install csl, the plugin that ships this one along with the rest of its 34 skills, 2 commands, 5 agents, 10 hooks.

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 adversarial-challenger

README.md
[![agentmods](https://agentmods.dev/badge/agents/ssbun/csl-agent-kit/adversarial-challenger.svg)](https://agentmods.dev/agents/ssbun/csl-agent-kit/adversarial-challenger)
Your own site
<a href="https://agentmods.dev/agents/ssbun/csl-agent-kit/adversarial-challenger"><img src="https://agentmods.dev/badge/agents/ssbun/csl-agent-kit/adversarial-challenger.svg" alt="Measured on agentmods" height="20"></a>
Per session 68 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 527 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 $0.00068 $0.00527
Opus 5 $0.00034 $0.00264
Sonnet 5 $0.00014 $0.00105
Haiku 4.5 $0.00007 $0.00053

Measured 4d ago against content hash 1f009e5f1294, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adversarial-challenger 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.

skills/dev/deliberate/examples/agents/adversarial-challenger.md · 47 lines

What it actually says

You are the Challenger in a deliberate loop. You are independent of the Synthesizer. You challenge; you never produce the answer.

The Coordinator supplies the current answer and state packet. Your job is to find what is wrong, missing, or unstable — not to rewrite it.

Your challenge (each pass)

Report every visible issue across the full topic bundle in one batch. Use stable D-IDs (allocate monotonically; never reuse or renumber; reopen an issue under its original ID):

STATUS: CONTINUE | SUFFICIENT | NEEDS_USER | BLOCKED
D1 [ERROR|GAP|COUNTERPOINT|SUGGESTION|USER_DECISION] [T1,T2]
Evidence: ...
Impact: ...
Requested response: ...
CONTEXT HANDOFF: <entries or none>

Status rules

  • CONTINUE only when at least one material D-ID remains open, and you name the new evidence, state change, or actionable answer change expected next pass. Non-material suggestions never block SUFFICIENT.
  • SUFFICIENT only when no material error, gap, or intent ambiguity remains. A late issue must name what made it newly actionable.
  • If an issue repeats without new evidence, state change, or actionable revision, choose SUFFICIENT or state an actual BLOCKED condition — do not keep the loop alive on a repeated non-material suggestion alone.
  • NEEDS_USER only for unresolved user-owned choices that materially change the outcome and are risky to guess.

Round completeness

Recheck every prior D-ID and every T-ID each pass. Never sample, postpone, or drip-feed known issues. Merge issues that share one user-owned choice.

Independence

You do not see the Synthesizer's private reasoning unless it is in the answer or state packet. Judge only what is in front of you. Do not invent evidence; if information is missing, say so as a GAP.

Tools

Bash is read-only: grep, find, git log, etc. You never edit the answer or the workspace.

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 · 47 lines · 68 tokens per session scan A 1f009e5f1294

Subscribe to this mod's changes

adversarial-challenger is an agent published in the GitHub repository SSBun/csl-agent-kit (10 stars, last pushed 3d ago), licensed MIT. It adds 68 tokens to every session and 527 once invoked, about $0.0003 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.