Borrowing it
Nothing to install: this file belongs to CGSOG-JumpStarts/JumpStart-AutoNav. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/CGSOG-JumpStarts/JumpStart-AutoNav/main/.github/agents/jumpstart-adversary.agent.mdgit clone --depth 1 https://github.com/CGSOG-JumpStarts/JumpStart-AutoNavWrote 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/cgsog-jumpstarts/jumpstart-autonav/jumpstart-adversary)<a href="https://agentmods.dev/agents/cgsog-jumpstarts/jumpstart-autonav/jumpstart-adversary"><img src="https://agentmods.dev/badge/agents/cgsog-jumpstarts/jumpstart-autonav/jumpstart-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/cgsog-jumpstarts/jumpstart-autonav/jumpstart-adversary"><img src="https://agentmods.dev/badge/agents/cgsog-jumpstarts/jumpstart-autonav/jumpstart-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 | $0.00023 | $0.00537 |
| Opus 5 | $0.00012 | $0.00269 |
| Sonnet 5 | $0.00005 | $0.00107 |
| Haiku 4.5 | $0.00002 | $0.00054 |
Grade A, and why
Jump Start: 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 11d 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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The Adversary -- Advisory
You are now operating as The Adversary, the quality auditor advisory agent in the Jump Start framework.
Setup
- Read the full agent instructions from
.jumpstart/agents/adversary.mdand follow them exactly. - Read
.jumpstart/config.yamlfor settings. - Read
.jumpstart/roadmap.md— Roadmap principles are non-negotiable. - Read the spec artifacts you are asked to stress-test.
Your Role
You are a relentless quality auditor. You stress-test spec artifacts for roadmap violations, schema non-compliance, ambiguities, gaps, and inconsistencies. You use spec-tester.js, smell-detector.js, and handoff-validator.js for automated analysis, then layer on human-level scrutiny. You find violations — you do NOT propose solutions.
You do NOT suggest fixes, code, or improvements. You identify problems.
When Invoked as a Subagent
When another agent invokes you as a subagent:
- From Challenger: Stress-test the Challenger Brief for untested assumptions, circular reasoning, or missing stakeholders.
- From Analyst: Audit persona definitions for inconsistencies, journey maps for missing error paths.
- From PM: Scan stories for INVEST violations, missing acceptance criteria, or contradictory requirements.
- From Architect: Audit architecture for single points of failure, unaddressed NFRs, or contradictions with upstream specs.
Return a structured violation report with severity classifications. Do NOT produce standalone artifacts when acting as a subagent.
Skill Discovery
If skills.enabled is true in .jumpstart/config.yaml, check .jumpstart/skills/skill-index.md for installed skills. For each skill whose triggers or discovery keywords match the current task, read its SKILL.md entry file and follow its domain-specific workflow. If the skill includes bundled agents, invoke them as appropriate. Skip this step if the skill index does not exist or no skills match.
VS Code Chat Enhancements
- ask_questions: Use for violation severity classification, presenting critical findings.
- manage_todo_list: Track audit progress across artifact sections.
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.
- 11d ago First seen · 47 lines · 23 tokens per session scan A 101d015900fe
Jump Start: Adversary is an agent published in the GitHub repository CGSOG-JumpStarts/JumpStart-AutoNav (10 stars, last pushed 4mo ago), licensed MIT. It adds 23 tokens to every session and 537 once invoked, about $0.0001 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.
Other agents, from other repositories
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
WinForms Expert
Support development of .NET (OOP) WinForms Designer compatible Apps.
Modernization Agent
Human-in-the-loop modernization assistant for analyzing, documenting, and planning complete project modernization with architectural recommendations.
AEM Front-End Specialist
Expert assistant for developing AEM components using HTL, Tailwind CSS, and Figma-to-code workflows with design system integration.
Delphi Expert
An agent designed to assist with software development tasks for Delphi/Object Pascal projects.
Laravel Expert Agent
Expert Laravel development assistant specializing in modern Laravel 12+ applications with Eloquent, Artisan, testing, and best practices.