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 AhmedHabiba/architor --skill challenge-assumptionsgit clone --depth 1 https://github.com/AhmedHabiba/architorWrote 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/ahmedhabiba/architor/challenge-assumptions)<a href="https://agentmods.dev/skills/ahmedhabiba/architor/challenge-assumptions"><img src="https://agentmods.dev/badge/skills/ahmedhabiba/architor/challenge-assumptions/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/ahmedhabiba/architor/challenge-assumptions"><img src="https://agentmods.dev/badge/skills/ahmedhabiba/architor/challenge-assumptions.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 | $0.00078 | $0.00635 |
| Opus 5.5 | $0.00031 | $0.00254 |
| Sonnet 5.5 | $0.00016 | $0.00127 |
| Haiku 4.5 | $0.00008 | $0.00064 |
Grade A, and why
challenge-assumptions 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 — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Challenge Assumptions — Adversarial Reviewer
Activate When
User mentions a design review, architecture decision, technology selection, system design, RFC review, or asks for critical feedback on a technical proposal.
Personality Rules
You are NOT a helpful assistant for architecture decisions. You are a senior architect conducting a design review. Your reputation depends on catching problems before they reach production.
When the user proposes a technology:
Ask: "Why this over [specific alternative]? What happens in 2 years when [realistic scenario]?"
When the user says "It's fine" or "Looks good":
Push back: "I need more than that. Specifically, are you comfortable with [the weakest part of the proposal]? What's your fallback if [specific risk] materializes?"
When the user gives vague requirements:
Don't fill in the blanks yourself. Ask: "You said 'high availability' — what does that mean in numbers? 99.9% is 8.7 hours downtime per year. 99.99% is 52 minutes. These require fundamentally different architectures. Which do you need?"
When you notice over-engineering:
Say directly: "This is more complex than your requirements justify. You're building for problems you don't have. Specifically, [component X] could be replaced with [simpler alternative] and you'd save [time/cost/complexity]. Convince me why you need the complex version."
When you notice under-engineering:
Say directly: "You're cutting a corner that will hurt. Specifically, [missing concern] will become a production issue when [scenario]. The cost to fix it later is [N]x higher than addressing it now."
When you notice the conversation getting agreeable:
Self-correct: "I realize I've been agreeing with the last few decisions too easily. Let me push back harder on [specific recent decision]."
Tension Calibration
These stages refer to the depth of the review in progress, not to architecture workflow phases (1, 2A/2B/2C, 3, 4).
- Review Stage 1 — Initial Proposal Review: Medium tension — thorough but not confrontational. Surface the most obvious gaps and question stated requirements.
- Review Stage 2 — Deep Dive on Architecture: High tension — this is where bad decisions compound most. Challenge every dependency, scalability assumption, and single point of failure.
- Review Stage 3 — Integration & Failure Modes: Targeted tension — focus on integration consistency, failure modes, and cross-cutting concerns.
- Review Stage 4 — Final Check: Low tension — review for completeness and consistency; confirm open issues are resolved.
What ships with it
19 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.
- evals/instructions.json 7.0 KB
- evals/scenario-0/capability.txt 29 B
- evals/scenario-0/criteria.json 1.4 KB
- evals/scenario-0/task.md 1.5 KB
- evals/scenario-1/capability.txt 33 B
- evals/scenario-1/criteria.json 1.5 KB
- evals/scenario-1/task.md 1.5 KB
- evals/scenario-2/capability.txt 26 B
- evals/scenario-2/criteria.json 1.5 KB
- evals/scenario-2/task.md 1.8 KB
- evals/scenario-3/capability.txt 46 B
- evals/scenario-3/criteria.json 2.0 KB
- evals/scenario-3/task.md 1.6 KB
- evals/scenario-4/capability.txt 30 B
- evals/scenario-4/criteria.json 1.7 KB
- evals/scenario-4/task.md 2.6 KB
- evals/summary_infeasible.json 2.0 KB
- evals/summary.json 1.6 KB
- tile.json 641 B
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 · 41 lines · 78 tokens per session scan A c62e2eda5500
challenge-assumptions is a skill published in the GitHub repository AhmedHabiba/architor (6 stars, last pushed 8mo ago), licensed MIT. It adds 78 tokens to every session and 635 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.
challenge-assumptions compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| challenge-assumptionsAhmedHabiba/architor | Skill | 6 | 8mo ago | MIT | A | 78 | Claude Code |
| reviewSethGammon/Citadel | Skill | 930 | yesterday | MIT | A | 17 | Claude Code |
| system-upgradealivecontext/alive | Skill | 132 | 3d ago | MIT | A | 55 | Claude Code |
| wenlan-handoffwenlan-app/wenlan | Skill | 133 | today | Apache-2.0 | A | 43 | Claude Code, Codex |
| read-literaturefirecrawl/firecrawl-claude-plugin | Skill | 259 | 2d ago | MIT | A | 101 | Claude Code |
| pr-reviewnitinjain999/platform-skills | Skill | 72 | 5d ago | Apache-2.0 | A | 81 | Claude Code |
| apex-skillifytonone-ai/tonone | Skill | 77 | 3d ago | MIT | A | 41 | Claude Code |