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 carbonshow/intent-fluid --skill expert-redteam-reviewgit clone --depth 1 https://github.com/carbonshow/intent-fluidWrote 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/carbonshow/intent-fluid/expert-redteam-review)<a href="https://agentmods.dev/skills/carbonshow/intent-fluid/expert-redteam-review"><img src="https://agentmods.dev/badge/skills/carbonshow/intent-fluid/expert-redteam-review/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/skills/carbonshow/intent-fluid/expert-redteam-review"><img src="https://agentmods.dev/badge/skills/carbonshow/intent-fluid/expert-redteam-review.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 | $0.00067 | $0.01087 |
| Opus 5.5 | $0.00027 | $0.00435 |
| Sonnet 5.5 | $0.00013 | $0.00217 |
| Haiku 4.5 | $0.00007 | $0.00109 |
Grade A, and why
expert-redteam-review 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 1mo 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 — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Expert Red-Team Review
Use this skill to improve decision quality for complex work. The main agent stays accountable for state, scope, synthesis, and final communication. Experts, red team, rebuttal, and judge roles exist to reveal blind spots; they do not replace the user's decision.
Core Rules
- Scale the workflow to the task. Do not run a full panel for trivial work.
- Keep expert initial judgments independent. Do not show one expert another expert's initial answer.
- Prefer 3-5 experts for serious reviews. More than 5 usually adds integration noise.
- Build vertical-domain experts from success criteria and failure modes; do not pretend a generic role is domain expertise.
- For vertical-domain L2/L3 work, propose the domain-specific panel and ask the user to confirm or adjust it before running the full review, unless the user has already approved the panel.
- Label important claims with evidence tags:
[source],[code],[test],[data],[inference],[engineering-judgment],[creative-judgment], or[unknown]. - Optimize for decision quality, not consensus. Preserve meaningful disagreement.
- The judge must arbitrate. A judge that only summarizes has failed.
- P0/P1 red-team findings require minimum fixes. P2 risks require explicit acceptance rationale.
- For one-way-door or high-blast-radius actions, stop at recommendation and ask for human approval before execution.
Level Selection
| Level | Use When | Shape |
|---|---|---|
| L0 Quick Check | The user wants a fast challenge or self-check. | Single-agent red-team review. |
| L1 Expert Review | 2-3 perspectives are useful, but formal adversarial review is unnecessary. | Small independent expert panel, no red team. |
| L2 Full Review | The user asks for expert panel/red team/judge or the decision has major uncertainty. | Expert panel, synthesis, red team, rebuttal, judge. |
| L3 Gated Review | The decision is hard to reverse or affects production, security, privacy, finance, compliance, or shared systems. | L2 plus explicit human gate before action. |
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 1mo ago First seen · 81 lines · 67 tokens per session scan A 89fa828a8081
expert-redteam-review is a skill published in the GitHub repository carbonshow/intent-fluid (6 stars, last pushed 1mo ago), licensed MIT. It adds 67 tokens to every session and 1,087 once invoked, about $0.0003 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 skills, from other repositories
swarm-advanced
Advanced swarm orchestration patterns for research, development, testing, and complex distributed workflows.
ai-agentic-workflows
Use this skill when the user says 'agentic workflow', 'agent loop', 'ReAct', 'plan and execute', 'planner executor', 'reflection loop', 'evaluator-optimizer', 'prompt chaining', 'LangGraph workflow'. Designs the control flow around an LLM: which workflow pattern to pick (chain, router, parallel, orchestrator-workers…
ticket-spec
Turn a raw Linear Triage ticket into a fully specified ai:agent-ready one by exploring the codebase. Use when specifying, refining, or promoting tickets, or when asked why a ticket isn't dispatchable.
ciel
Ciel — self-improving, self-researching orchestration intelligence. A living skill graph that routes, acquires, integrates, and evolves capabilities across any skill-aware agent runtime.
Exception Documentation
The template for documenting a deliberate, approved deviation from an infrastructure invariant — rationale, compensating control, approver, and expiry — instead of a silent workaround.
design-token-architecture
CIEL's framework for three-layer design token systems (primitive→semantic→component) with CSS variables and naming conventions. Advisory only.