agentmods.dev

research-redteam-loop

research-redteam-loop is a skill for Claude Code from chf3198/copilot-governance. It costs 55 tokens per session (1,180 once invoked), scanned A, original, MIT.

A repeated research and review process that produces a deliverable, has another model challenge it, and revises it until it reaches a required score or the iteration limit.

In plain words
What is it for?
Use it for issue-specific research, corpus-overlap checks across open and closed issues, adversarial review, scoring, and repeated rework.
Why use it?
It catches missing evidence, overlap with other issues, and weak conclusions before a research result is accepted.

Skill for Claude Code

Written for Claude Code: disable-model-invocation in frontmatter. Also seen: reads .claude/ paths; positional $N argument.

not rated 1repo 1mo ago A scan 55 tokens original MIT

Good fit Use it for issue-specific research, corpus-overlap checks across open and closed issues, adversarial review, scoring, and repeated rework.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/chf3198/copilot-governance/research-redteam-loop
View source ↗ chf3198/copilot-governance
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.

Any agent
npx skills add chf3198/copilot-governance --skill research-redteam-loop
Clone the repo
git clone --depth 1 https://github.com/chf3198/copilot-governance

Made for: Claude Code.

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 research-redteam-loop

README.md
[![agentmods](https://agentmods.dev/badge/skills/chf3198/copilot-governance/research-redteam-loop.svg)](https://agentmods.dev/skills/chf3198/copilot-governance/research-redteam-loop)
Your own site
<a href="https://agentmods.dev/skills/chf3198/copilot-governance/research-redteam-loop"><img src="https://agentmods.dev/badge/skills/chf3198/copilot-governance/research-redteam-loop.svg" alt="Measured on agentmods" height="20"></a>
Per session 55 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,180 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00055 $0.01180
Opus 5 $0.00028 $0.00590
Sonnet 5 $0.00011 $0.00236
Haiku 4.5 $0.00006 $0.00118

Measured 6d ago against content hash 2f2e90f1972f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

research-redteam-loop 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 6d 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/research-redteam-loop/SKILL.md · 94 lines

How it starts

The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Research Red-Team Loop Skill

Purpose

Bundles the proven research→adversarial-review→rework→consensus cycle (validated in #2710: 78→92→91→94 ACCEPT) into a single invocation. Fleet/free-cloud only. Cross-family invariant: reviewer MUST NOT be Anthropic/Claude family.

Invocation

/research-redteam-loop #N [--gate 93] [--web] [--cap 5]

Arguments: --gate N = minimum acceptance score (default 93). --web = inject web search evidence (default on; degrades gracefully). --cap N = max iterations.

Loop Protocol

  1. Produce/refresh the research deliverable for issue #N (read existing body or linked wiki page). 1a. Corpus-overlap stage (MANDATORY before ACCEPT — #2801). Run node scripts/global/corpus-overlap-scan.js --scope-file <deliverable> --json. It reuses the #2617 tokenizer but scans open AND closed issues (the near-miss #2518 was a closed epic) via overlap-coefficient. Embed buildOverlapBlock (related_tickets: + overlap_decision:) in the deliverable and inject reviewerOverlapFragment into the reviewer prompt so the red-team validates the boundary (overlap vs conflict vs redundancy vs gap). ACCEPT is blocked when shouldBlockAccept(scan, deliverable) is true (overlap present, no overlap_decision:). Distinct from #2617/Epic #2623 (ticket-CREATION stage).
  2. Web augmentation: fetch 3+ relevant references via WebSearch; inject into reviewer prompt. On search failure, log web-search-unavailable to ~/.megingjord/incidents.jsonl and continue without web context.
  3. Dispatch review (G3 fleet-first) — canonical CLI is cascade-dispatch.js (#2858 / Epic #2926 D1):
    • node scripts/global/cascade-dispatch.js --prompt "<review prompt>" --model qwen2.5-coder:32b [--json] — orchestrates fleet → free-cloud fallback → escalation; timeouts per timeout-policy.json. It already emits [cascade] escalate→<tier> to stderr on escalation.
    • On a fleet availability failure it fails over to the $0 free-cloud chain automatically (#2619/#2621); premium is reached only on a free-cloud capability failure (Epic #2926 D4).
    • fleet-red-team-dispatch.js is a LIBRARY (no CLI) — do NOT invoke it directly; it redirects here.
    • Cross-family invariant enforced at every tier (no Anthropic model).
  4. Score check: if score > gate and shouldBlockAccept is false → post ACCEPT and stop. If overlap blocks (no overlap_decision:), record the boundary and re-review. Else → post ITER comment and proceed to rework.
  5. Manager rework: revise deliverable using reviewer findings + web context.
  6. Repeat steps 2–5 until ACCEPT or cap exhausted → post REJECT.

Read the full file on GitHub · 94 lines

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. 6d ago First seen · 94 lines · 55 tokens per session scan A 2f2e90f1972f

Subscribe to this mod's changes

research-redteam-loop is a skill published in the GitHub repository chf3198/copilot-governance (1 stars, last pushed 1mo ago), licensed MIT. It adds 55 tokens to every session and 1,180 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.

Research
Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens