ai-agent-architecture-reviewer
Skill7 starsChanged 8 days agoMITView on GitHub
Review AI agent systems for architecture flaws, planning failures, memory risks, tool abuse, MCP security, hallucination and verification gaps, prompt injection, multi-agent coordination failures, observability gaps, reliability risks, cost explosions, scalability limits, and production readiness. Use for single agents, multi-agent systems, copilots, workflow agents, browser agents, coding agents, research agents, and operations agents.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- letscooktech-studio/ai-engineering-arsenalword for word, first seen Nov 8, 2026
Other versions1 project uses a version of this.
Word for word · 1
- letscooktech-studio/ai-engineering-arsenal★ 7100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository17 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agentic-skill-pack-integratorSkill
- ai-search-optimizerSkill
- competitor-analyzerSkill
- cost-explosion-detectorSkill
- cto-operating-systemSkill
- database-architectSkill
- nextjs-production-architecture-reviewerSkill
- nvidia-accelerated-ai-systems-operatorSkill
- production-ai-saas-builderSkill
- production-risk-labSkill
- security-auditorSkill
- seo-auditorSkill
- startup-validatorSkill
- supabase-production-auditorSkill
- system-architectSkill
- technical-debt-hunterSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 29
Do not approve an agent that can take external actions without permission boundaries, tool validation, logging, failure handling, and …
- Distinctive: line 30
Do not trust model self-critique as verification unless it is backed by external evidence, tests, or deterministic checks.
- Distinctive: line 31
Treat prompt injection, tool abuse, secret exposure, and uncontrolled cost fan-out as high-risk until mitigated.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AI Agent Architecture Reviewer
Developed by the LetsCookTech Open Source Team.
Apply the shared OutcomeLab skill standard for evidence, verification, risk, benchmarking, and improvement.
Review AI agents as production systems with planning, memory, tools, permissions, verification, observability, cost, and failure behavior.
Workflow
- Classify agent type, autonomy level, execution pattern, business criticality, users, permissions, tools, model providers, memory stores, and deployment environment.
- Map architecture: agent boundaries, decision layers, execution layers, tool registry, memory flow, orchestration, queues, fallback systems, escalation paths, and dependency graph.
- Audit planning, memory, tooling, verification, multi-agent coordination, prompt-injection defenses, tool-output validation, reliability, observability, scalability, and cost controls.
- Red-team misuse: data leakage, prompt injection, indirect prompt injection, tool abuse, secret exposure, privilege escalation, infinite loops, agent drift, hallucinated actions, and spend explosions.
- Score architecture, planning, memory, tooling, verification, security, reliability, scalability, cost efficiency, observability, production readiness, and maintainability.
- Produce evidence-backed findings, mitigation design, verification tests, operational controls, 30-day roadmap, and 90-day roadmap.
Output contract
Return executive summary, critical risks, architecture findings, planning findings, memory findings, tooling findings, verification findings, security findings, reliability findings, scalability findings, cost findings, observability findings, production readiness findings, recommended fixes, 30-day roadmap, 90-day roadmap, final assessment, and confidence/risk scores.
Quality gates
- Do not approve an agent that can take external actions without permission boundaries, tool validation, logging, failure handling, and escalation.
- Do not trust model self-critique as verification unless it is backed by external evidence, tests, or deterministic checks.
- Treat prompt injection, tool abuse, secret exposure, and uncontrolled cost fan-out as high-risk until mitigated.
- Include a safe stop condition for autonomous or long-running agents.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.