ai-agent-governance-and-limits
Skill31 starsChanged 18 days agoMITView on GitHub
Use when defining agent budgets, step limits, reversibility, blast-radius controls, kill switches, and governance policy for agentic AI systems.
In this repository172 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- 00-meta-initializationSkill
- chwezi-dev-engineSkill
- ai-agent-commercial-operationsSkill
- ai-agent-compliance-controlsSkill
- ai-agent-multi-agent-coordinationSkill
- ai-agent-observability-evaluationSkill
- ai-agent-runtime-architectureSkill
- ai-agent-safety-and-red-teamSkill
- ai-agent-tooling-and-hitlSkill
- ai-analyticsSkill
- ai-app-architectureSkill
- ai-cost-and-meteringSkill
- ai-economic-value-engineSkill
- ai-evaluationSkill
- ai-feature-rollout-and-experimentationSkill
- ai-feature-specSkill
- ai-incident-responseSkill
- ai-llm-integrationSkill
- ai-model-gatewaySkill
- ai-observability-and-debuggingSkill
- ai-opportunity-canvasSkill
- ai-prompt-engineeringSkill
- ai-rag-patternsSkill
- ai-securitySkill
- ai-web-appsSkill
- coding-agent-optimizationSkill
- openai-agents-sdkSkill
- android-developmentSkill
- api-design-firstSkill
- distributed-systems-patternsSkill
- ecommerce-platform-audit-requirementsSkill
- graphql-patternsSkill
- microservices-architectureSkill
- system-architecture-designSkill
- validation-contractSkill
- database-design-engineeringSkill
- database-reliabilitySkill
- mysql-engineeringSkill
- postgresql-engineeringSkill
- cicd-pipelinesSkill
- cloud-architectureSkill
- deployment-release-engineeringSkill
- docker-developmentSkill
- infrastructure-as-codeSkill
- kubernetes-platformSkill
- observability-monitoringSkill
- reliability-engineeringSkill
- execution-plan-scriptsSkill
- accounting-engineSkill
- accounting-finance-controllerSkill
- electronic-fiscal-taxingSkill
- multicurrency-and-fxSkill
- avalonia-desktop-developmentSkill
- frontend-architectureSkill
- frontend-performanceSkill
- image-compressionSkill
- nextjs-app-routerSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is 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.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 35
Read this SKILL.md first.
- Distinctive: line 42
Keep routing explicit: name which reference files were used when the work depends on absorbed material.
- Distinctive: line 44
Prefer concrete contracts, checklists, tables, schemas, runbooks, and decision records over broad summaries.
- Distinctive: line 81
Read and search are required.
≈ “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 Governance And Limits
Acknowledgement: Shared by Peter Bamuhigire, techguypeter.com, +256 784 464178.
<!-- dual-compat-start -->Use When
- Set step, wallclock, token, model, and cost budgets for agents.
- Define reversibility, blast-radius limits, and safe failure behavior.
- Connect runtime controls to policy, tenant limits, and operational review.
Do Not Use When
- The work is not AI-specific or agentic-AI-specific.
- A narrower retained AI parent skill fits the request better.
Required Inputs
- Product, tenant, user, data, risk, and operational context relevant to the AI workflow.
- Target artifact: design, implementation plan, audit, test strategy, UX flow, commercial policy, or runbook.
- Constraints from security, privacy, reliability, billing, support, and compliance stakeholders when relevant.
Workflow
- Read this SKILL.md first.
- Load references/routing.md to select the absorbed child reference that matches the task.
- Load only the selected child reference files needed for the current request.
- Produce execution-oriented output with assumptions, risks, evidence, and next actions where relevant.
Quality Standards
- Keep routing explicit: name which reference files were used when the work depends on absorbed material.
- Preserve tenant isolation, auditability, cost controls, safety gates, and operational evidence when they matter.
- Prefer concrete contracts, checklists, tables, schemas, runbooks, and decision records over broad summaries.
Anti-Patterns
- Loading every absorbed reference by default.
- Treating AI-specific billing, compliance, safety, or UX concerns as generic SaaS work without checking AI failure modes.
- Hiding retired skill names; old slugs must remain discoverable through references/routing.md.
Outputs
- A concrete deliverable matched to the request: architecture, implementation plan, audit, policy, runbook, UX flow, test strategy, or operating model.
- The selected consolidated reference files and any assumptions, risks, evidence requirements, or follow-up actions that affect execution.
References
- references/routing.md maps retired child skill slugs to their consolidated reference folders.
- Load references/agent-governance-and-oversight.md when deciding code-versus-model split, deferral to humans, safeguard components, agent registry, readiness, or NIST AI RMF evidence mapping.
Consolidated Child References
- Load references/routing.md to map retired AI child skill slugs to their reference modules.
Evidence Produced
| Category | Artifact | Format | Example |
| Security | Agent limits and reversibility register | Markdown table | step budget, blast radius, approval tier, kill path, and drill result |
<!-- dual-compat-end -->Inputs
| Artefact | Required? | Purpose |
| Agent task and tool catalogue | yes | Identify actions and costs |
| Step, token, time, money, and blast-radius policy | yes | Define limits |
| Escalation and termination policy | yes | Stop unsafe runs |
Capability contract
Read and search are required. Budget enforcement tests may execute only in controlled environments. Changing production limits or kill switches requires explicit operational authority.
Degraded mode
Fallback without telemetry or enforcement access: produce the limit policy and gap register; do not claim limits are enforced.
Decision rules
| Limit event | Runtime response | Failure avoided |
| Soft threshold reached | Warn, checkpoint, or degrade | Abrupt failure |
| Hard cost/step/time threshold reached | Stop safely and preserve state | Runaway execution |
| Blast-radius or policy boundary reached | Deny and escalate | Unsafe autonomy |
Domain anti-patterns
- Limits documented only in prompts. Fix: enforce outside the model.
- One global budget for every tenant and task. Fix: scope by risk and entitlement.
- Retrying after a hard limit. Fix: require continuation authority.
- Kill switch without state preservation. Fix: checkpoint and record termination reason.
- Measuring attempted steps as completed value. Fix: separate attempts and success.
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.