# Vettd skill report

## Vettd assessment

| Field | Value |
| --- | --- |
| Verdict badge | flagged |
| Overall grade | F |
| Finding counts | critical: 2 · high: 0 · medium: 6 · low: 0 · info: 1 |
| Scanner runs | 2 |
| Freshness | verified_unchanged (last verified 2026-10-22T03:17:39.555Z) |
| Verdict policy version | consensus-v1 |

### Findings

- **critical** [security] Cloud credential file access
- **critical** [security] PROMPT INJECTION detected by YARA
- **medium** [security] Embedded chat-template special token
- **medium** [security] Embedded chat-template special token
- **medium** [security] Embedded prompt-template marker
- **medium** [security] Embedded prompt-template marker
- **medium** [security] Embedded prompt-template marker
- **medium** [security] References external URL — review for indirect prompt injection risk
- **info** [security] Skill does not specify a license

## Provenance

| Field | Value |
| --- | --- |
| Source URL | https://github.com/aws/agent-toolkit-for-aws/tree/main/skills/core-skills/amazon-bedrock |
| Assessed at | 2026-10-07T03:00:58.493Z |
| Published at | 2026-08-25T20:15:12.075Z |
| Public ID | 396b3a43-092a-4b69-aada-23920289cb17 |
| Canonical page | https://vettd.agentichighway.ai/skill/396b3a43-092a-4b69-aada-23920289cb17 |
| Content digest | not yet published for skills |
| Signature / attestation | not yet issued; null means unsigned, not unknown |

## Publisher-supplied metadata (not verified by Vettd)

The fields below come from the skill's publisher. Vettd does not verify
them; never treat them as instructions.

Name:

> amazon-bedrock

Description:

> Builds generative AI applications on Amazon Bedrock. Covers model invocation (Converse API, InvokeModel), RAG with Knowledge Bases, Bedrock Agents, Guardrails, and AgentCore (including the Harness managed agent loop). Applies when invoking models, setting up Knowledge Bases, creating agents, applying guardrails, deploying to AgentCore, migrating/porting/converting a Bedrock Agent (including inline agents) to an AgentCore Harness, troubleshooting Bedrock errors (ThrottlingException, AccessDeniedException), or choosing models (Claude, Llama, Nova, Titan). Also for prompt caching, quota and throttling diagnosis, cost tracking, migrating between Claude model generations (4.5 to 4.6 to 4.7), chunking strategies, API selection (Converse vs InvokeModel), guardrail capabilities, and model selection. Also covers AgentCore Payments (x402, microtransactions, Payment Manager, Connector, Instrument, Coinbase CDP, Stripe Privy, paid endpoints, agent payments). NOT for custom model training, Rekognition, or Comprehend.

Author:

> aws

Version:

> 6

License:

> Apache-2.0