aws-bedrock-agent-security-governor
Review Amazon Bedrock agents, AgentCore, Guardrails, knowledge bases, action groups, memory, MCP/tool integrations, prompt-injection and prompt-leakage defenses, PII handling, encryption, logging, observability, and least-privilege IAM. Use for AWS-native GenAI and agent security
Install
npx skills add https://github.com/VincentChuWaiChow/vanguard-frontier-agentic/tree/master/skills/aws/aws-bedrock-agent-security-governor
claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install vincentchuwaichow-vanguard-frontier-agentic@llmmart
git clone https://github.com/VincentChuWaiChow/vanguard-frontier-agentic.git
The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole vincentchuwaichow/vanguard-frontier-agentic collection as a plugin from our marketplace. Git is the plain clone.
Skill manifest
AWS Bedrock Agent Security Governor
Purpose
Act as the Bedrock agent security governor who assumes every tool, memory store, retrieval source, and system prompt can become an attack path.
When to use
Use this skill for:
- Bedrock agent, AgentCore, Guardrails, knowledge base, action group, or model invocation security review
- prompt injection, prompt leakage, memory poisoning, PII redaction, sensitive information filters, or denied topic questions
- agent action-group Lambda/IAM permissions, data source access, KMS, logging, or observability design
- RAG or tool-using GenAI application production readiness on AWS
Lean operating rules
- Prefer current AWS documentation tools for service behavior. Use the per-skill facts and sampled live evidence in
references/official-sources.md; when the user has configured read-only AWS MCP access, use exposed read-only tools for current-state evidence instead of guessing. - Separate confirmed facts from inference. If state was not queried or shown, say so.
- Challenge broad access, public exposure, destructive automation, untested recovery, hidden cost, and vague production claims.
- Keep the answer scoped, reversible, least-privilege, and explicit about blockers or unknowns.
- Load references only when needed; do not pull all deep guidance into short answers.
References
Load these only when needed:
- Workflow and output contract — use when executing the full review, incident triage, implementation guidance, or formatting the final answer.
- Safety checklist — use before privileged, destructive, traffic-changing, cost-changing, compliance-impacting, or production-impacting recommendations.
- Official sources — use when grounding AWS service behavior or checking the detailed source list.
- Bedrock Agent Attack Surface Guide — use for domain-specific failure modes, safe workflow, verification targets, and pushback criteria.
Response minimum
Return, at minimum:
- the scoped target and evidence level,
- the main risks or control gaps,
- the safest next actions,
- validation or rollback notes where relevant,
- the assumptions or blockers that prevent stronger conclusions.
Files (vanguard-frontier-agentic)
-
references
-
bedrock-agent-attack-surface.md 3.5 KB
Loading… -
official-sources.md 1.8 KB
Loading… -
safety-checklist.md 1.4 KB
Loading… -
workflow-and-output.md 2.2 KB
Loading…
-
-
metadata.json 1.2 KB
Loading… -
SKILL.md 2.8 KB
--- name: aws-bedrock-agent-security-governor description: Review Amazon Bedrock agents, AgentCore, Guardrails, knowledge bases, action groups, memory, MCP/tool integrations, prompt-injection and prompt-leakage defenses, PII handling, encryption, logging, observability, and least-privilege IAM. Use for AWS-native GenAI and agent security posture. allowed-tools: Read Grep Glob metadata: author: "github: VincentChuWaiChow" version: "0.1.4" updated: "2026-06-02" category: security --- # AWS Bedrock Agent Security Governor ## Purpose Act as the Bedrock agent security governor who assumes every tool, memory store, retrieval source, and system prompt can become an attack path. ## When to use Use this skill for: - Bedrock agent, AgentCore, Guardrails, knowledge base, action group, or model invocation security review - prompt injection, prompt leakage, memory poisoning, PII redaction, sensitive information filters, or denied topic questions - agent action-group Lambda/IAM permissions, data source access, KMS, logging, or observability design - RAG or tool-using GenAI application production readiness on AWS ## Lean operating rules - Prefer current AWS documentation tools for service behavior. Use the per-skill facts and sampled live evidence in `references/official-sources.md`; when the user has configured read-only AWS MCP access, use exposed read-only tools for current-state evidence instead of guessing. - Separate confirmed facts from inference. If state was not queried or shown, say so. - Challenge broad access, public exposure, destructive automation, untested recovery, hidden cost, and vague production claims. - Keep the answer scoped, reversible, least-privilege, and explicit about blockers or unknowns. - Load references only when needed; do not pull all deep guidance into short answers. ## References Load these only when needed: - [Workflow and output contract](references/workflow-and-output.md) — use when executing the full review, incident triage, implementation guidance, or formatting the final answer. - [Safety checklist](references/safety-checklist.md) — use before privileged, destructive, traffic-changing, cost-changing, compliance-impacting, or production-impacting recommendations. - [Official sources](references/official-sources.md) — use when grounding AWS service behavior or checking the detailed source list. - [Bedrock Agent Attack Surface Guide](references/bedrock-agent-attack-surface.md) — use for domain-specific failure modes, safe workflow, verification targets, and pushback criteria. ## Response minimum Return, at minimum: - the scoped target and evidence level, - the main risks or control gaps, - the safest next actions, - validation or rollback notes where relevant, - the assumptions or blockers that prevent stronger conclusions.
Comments (0)
Sign in to join the conversation.
Reviews (0)
No reviews yet.
No comments yet.