Bedrock Multi-Model
NPM · BEDROCK-MULTI-MODEL-MCP
Ask, compare and judge 35 Bedrock text models, generate and edit images, plus video and embeddings.
Available components
Last scanned (58 minutes ago)
Scores higher than 83% of npm servers
Category AI and Machine Learning
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security98
- No malware found by supply-chain analysis.Pass
- Known CVEs were checked across the 119 of 120 dependencies we could resolve, so this covers what we could see, not the whole tree.Partial
- No install/post-install scripts declared.Pass
- Dependency health was assessed across the 119 of 120 dependencies we could resolve, so this covers what we could see, not the whole tree. View diagnostics → Partial
Provenance & Transparency100
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Cryptographically verified build provenance (signed, bound to Sagargupta16/bedrock-multi-model-mcp). View diagnostics → Pass
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 1 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability72
- 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Pass
- AI-judged instruction clarity (excellent).Pass
- Context-footprint check failed: tool/resource definitions use about 4906 tokens (~446/item across 11 items; 11 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management70
- Stability observed for 21 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage100
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 100% of tool parameters carry a description.Pass
Tool Safety75
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- 0 of 3 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation; "bedrock_ask" implies "send" and declares readOnlyHint instead, contradicting what its own name says it does. See how to fix → Fail
- An AI judge read all 11 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Checks we ran (16)
- Malware scan
- Known CVEs
- Install scripts
- Dependency health
- Source repository
- Provenance
- License
- Maintenance
- Security disclosure
- Schema quality
- Stability
- Tool coverage
- Injection markers
- Destructive annotations
- Judged manipulation
- MCP protocol
Does not apply to this package (2)
- Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
- Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
The score does not measure everything. What the score does not measure →
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 22 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 67 to 70. That category is still filling its 30-day observation window: 20 days of observed history at the previous scan, 21 at this one. The score rises as the window fills, whether or not the server changes.
- 21 Nov 26 +11
- Known CVEs: unverified → partial ▲ security
- Dependency health: unverified → partial ▲ functional
- 20 Nov 26 −8
- Known CVEs: partial → unverified ▼ security
- Known CVEs: partial → pass ▲ security
- Schema quality: 286 → 446 ▼ functional
- Dependency health: partial → unverified ▼ functional
- First check of Schema quality: 100 functional
- Dependency health: partial → 0.89 functional
- Package version: 0.8.0 → 0.9.0 functional
- 18 Nov 26 +1
- Known CVEs: pass → partial ▼ security
- Dependency health: 0.89 → partial functional
- 15 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 43 to 47. That category is still filling its 30-day observation window: 13 days of observed history at the previous scan, 14 at this one. The score rises as the window fills, whether or not the server changes.
- 13 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 37 to 40. That category is still filling its 30-day observation window: 11 days of observed history at the previous scan, 12 at this one. The score rises as the window fills, whether or not the server changes.
- 11 Nov 26 +16
- Malware scan: unverified → pass ▲ security
- 10 Nov 26 −15
- Malware scan: pass → unverified ▼ security
- Package version: 0.7.1 → 0.8.0 functional
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 22 Nov 2026 · Analysed npm/bedrock-multi-model-mcp@0.9.0
Provenance Verified
A signed build attestation was found and verified, binding this exact artifact to the source repository it claims to come from.
| Result | Verified |
|---|---|
| Ecosystem | npm |
| Reason | Verified |
| Discovered via | Registry attestation endpoint |
| Source repo | Sagargupta16/bedrock-multi-model-mcp |
| Certificate issuer | https://token.actions.githubusercontent.com |
| Certificate SAN | https://github.com/Sagargupta16/bedrock-multi-model-mcp/.github/workflows/release.yml@refs/tags/v0.9.0 |
| Rekor log index | 4017513076 |
| Predicate type | SLSA build provenance https://slsa.dev/provenance/v1 |
| Subject digest | sha512:c77122620109d65972a4d06910d37b6d17087deb929e8d9ff9ee02e95d7c09aae872165c90b5bba21fcad7ed02eef848fff92be70df786e53de439400 |
Background: How many MCP packages publish verified provenance →
Dependencies 119 packages
| Packages resolved | 119 |
|---|---|
| Stale | 31 |
| Tree resolution | Partial |
The dependency tree was only partially resolved, so these counts may be incomplete.
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
bedrock_ask ~821
Send a prompt to any AWS Bedrock model via the Converse API. Supports Claude, GPT, Grok, Kimi, Gemma, Mistral, Nova, Qwen, DeepSeek, and more. Use short aliases (e.g. 'claude-sonnet', 'nova-pro') or full model IDs, or pass level instead of model to use the first model of that level. Pass schema to get JSON that matches it, and effort to set the thinking depth. files: Optional local text files to include in the prompt, read only from under the root folder (env BEDROCK_MCP_FILES_ROOT, default the server's working folder unless that is a home folder or a drive root): up to 10 files, 256 KB each and 1 MB in total, secret-looking files such as .env and keys and network paths are refused, and the content is sent to the model provider. A refused file fails the call before any model is billed. Optional conversation_id turns on a follow-up conversation: earlier turns of the same id are sent along as context. It is kept in memory only, forgotten 30 minutes after the last turn or when the server restarts, holds at most 20 turns, and the id is 1 to 64 letters, digits, "-" or "_". Attached file contents are kept with their turn and sent again on each follow-up, to whichever model that follow-up names. The reply header shows the tokens and an estimated cost from list prices.
| Name | Type | Req | Description |
|---|---|---|---|
| conversation_id | string | – | Follow-up conversation id (1-64 letters, digits, '-' or '_'); omit for a one-off call. See the tool description. |
| effort | string | – | Thinking depth and token spend for models that take it (Claude except Haiku 4.5, GPT-6.x, Grok 4.7, Kimi K3, Gemma 4 up to xhigh): low, medium, high, xhigh, max. Claude Opus 5.5 and Haiku 5.5 default… |
| files | array | – | Local text file paths, relative to the files root (BEDROCK_MCP_FILES_ROOT or the server's working folder). Each file is appended to the prompt and its content is sent to the model provider. |
| level | string | – | Use the first model of a level instead of model: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT… |
| max_tokens | integer | – | Maximum tokens to generate |
| model | string | – | Model alias (e.g. 'claude-opus', 'kimi', 'mistral-large', 'nova-pro', 'deepseek') or full Bedrock model ID. Required unless level is given; model and level together is an error. |
| prompt | string | yes | The user prompt to send to the model |
| schema | object | – | JSON schema the reply must match (structured output). Object schemas need additionalProperties: false. Supported by Claude, GPT-6.x, GPT-OSS, Grok 4.7, Kimi K3, Gemma 4, Mistral Large 3, Ministral 3,… |
| system | string | – | Optional system prompt |
| temperature | number | – | Sampling temperature (0-1, default 0.7). Ignored by models that don't support it. |
No output schema declared.
No examples provided.
bedrock_compare ~698
Send the same prompt to multiple Bedrock models and compare responses side by side. Useful for evaluating model quality, speed, and style differences. Pass level instead of models for a default set of three: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT-6.2 Astra, Grok 4.7); models wins when both are given. files: Optional local text files to include in the prompt, read only from under the root folder (env BEDROCK_MCP_FILES_ROOT, default the server's working folder unless that is a home folder or a drive root): up to 10 files, 256 KB each and 1 MB in total, secret-looking files such as .env and keys and network paths are refused, and the content is sent to the model provider. A refused file fails the call before any model is billed. Each answer shows its tokens and an estimated cost from list prices, with the total at the end.
| Name | Type | Req | Description |
|---|---|---|---|
| effort | string | – | Thinking depth and token spend for models that take it (Claude except Haiku 4.5, GPT-6.x, Grok 4.7, Kimi K3, Gemma 4 up to xhigh): low, medium, high, xhigh, max. Claude Opus 5.5 and Haiku 5.5 default… |
| files | array | – | Local text file paths, relative to the files root (BEDROCK_MCP_FILES_ROOT or the server's working folder). Each file is appended to the prompt and its content is sent to the model provider. |
| level | string | – | Default models when models is absent: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT-6.2 Astra,… |
| max_tokens | integer | – | Maximum tokens per model response |
| models | array | – | List of 2-5 model aliases or IDs to compare (e.g. ['kimi', 'nova-pro', 'mistral-large']). Required unless level is given. |
| prompt | string | yes | The prompt to send to all models |
| schema | object | – | JSON schema the reply must match (structured output). Object schemas need additionalProperties: false. Supported by Claude, GPT-6.x, GPT-OSS, Grok 4.7, Kimi K3, Gemma 4, Mistral Large 3, Ministral 3,… |
| system | string | – | Optional system prompt sent to all models |
| temperature | number | – | Sampling temperature for all models |
No output schema declared.
No examples provided.
bedrock_doctor ~178
Check which Bedrock text models work on this AWS account right now. Sends one small billed call (a one-word prompt, at most 32 output tokens) to each model: every registry text model (35) by default, or only the models passed. Returns a table with one status per model: ok (answered), throttled (rate or quota limit right now; try again later), denied (no model access or not authorized on this account), invalid (unknown model id, not in the registry, or not offered in this region), error (anything else; see the note).
| Name | Type | Req | Description |
|---|---|---|---|
| concurrency | integer | – | Models checked at the same time, 1-8 (default 4) |
| models | array | – | Model aliases or IDs to check (default every text model in the registry); a repeated model is checked once |
No output schema declared.
No examples provided.
bedrock_edit_image ~712
Edit or transform a local image with Stability AI Image Services on AWS Bedrock: fill or erase a masked area, extend the canvas, swap or recolor an object found by a prompt, remove the background, upscale, draw from a sketch or structure, or apply a style. Saves the result as a PNG, like bedrock_generate_image.
| Name | Type | Req | Description |
|---|---|---|---|
| aspect_ratio | string | – | style-guide: output aspect ratio (default 1:1) |
| change_strength | number | – | style-transfer: how much the input may change (default 0.9) |
| composition_fidelity | number | – | style-transfer: how closely the output keeps the input's composition (default 0.9) |
| control_strength | number | – | control-sketch, control-structure: how closely the output follows the input (default 0.7) |
| creativity | number | – | upscale-conservative and upscale-creative (0.1-0.5), outpaint (0.1-1): how much new detail to invent |
| down | integer | – | outpaint: pixels to add at the bottom |
| fidelity | number | – | style-guide: how closely the output follows the input image's style (default 0.5) |
| grow_mask | integer | – | inpaint, erase, search-replace, search-recolor: pixels to grow and soften the mask edge (default 5) |
| image_path | string | yes | Input image (PNG, JPEG or WebP), absolute or relative to cwd. For style-transfer, the image to restyle. |
| left | integer | – | outpaint: pixels to add on the left |
| mask_path | string | – | inpaint, erase: black-and-white mask image, white where the image should change. Without it, the input image's alpha channel is used as the mask. |
| negative_prompt | string | – | What to keep out of the output |
| operation | string | yes | What to do: inpaint (needs prompt); outpaint (needs one of left/right/up/down); erase; remove-background; search-replace (needs prompt, search_prompt); search-recolor (needs prompt, select_prompt); u… |
| output_dir | string | – | Directory to save the image, as in bedrock_generate_image (BEDROCK_MCP_OUTPUT_DIR, cwd or ~/bedrock-images) |
| output_format | string | – | Result file format (default png; webp for the three upscale operations, whose PNG is over Bedrock's 16 MB response limit) |
| prompt | string | – | What the output should show |
| right | integer | – | outpaint: pixels to add on the right |
| search_prompt | string | – | search-replace: the object to find and replace |
| seed | integer | – | Random seed for reproducible results |
| select_prompt | string | – | search-recolor: the object to find and recolor |
| style_image_path | string | – | style-transfer: the image whose style is applied |
| style_preset | string | – | Style preset, for the operations that take one |
| style_strength | number | – | style-transfer: how strongly the style image applies (default 1) |
| up | integer | – | outpaint: pixels to add on top |
No output schema declared.
No examples provided.
bedrock_embed_similarity ~213
Embed 2 or more texts with an AWS Bedrock embedding model (Amazon Titan, Cohere) and return a cosine-similarity matrix. Scores range 0 (unrelated) to 1 (identical meaning). Useful for semantic search, deduplication, and clustering. Pass images (local JPEG or PNG paths) beside or instead of texts to compare images with each other and with texts; that needs a multimodal model ('titan-multimodal' or 'cohere'). At least 2 inputs in total.
| Name | Type | Req | Description |
|---|---|---|---|
| images | array | – | Local JPEG or PNG paths (absolute or relative to cwd) to embed beside the texts; rows and columns show the file name |
| model | string | – | Embedding model alias: 'titan-v2' (default), 'titan-multimodal', 'cohere' (Embed v4), or a full model ID |
| texts | array | – | List of up to 50 texts to embed and compare pairwise |
No output schema declared.
No examples provided.
bedrock_generate_image ~338
Generate an image from a text prompt using AWS Bedrock image models (Stable Image Ultra, Stable Diffusion 3.5 Large, Stable Image Core). Saves the image to the current working directory by default (or output_dir if passed, or BEDROCK_MCP_OUTPUT_DIR env var, falling back to ~/bedrock-images/ if cwd is not writable).
| Name | Type | Req | Description |
|---|---|---|---|
| height | number | – | Desired height (default 1024). See `width` - only the resulting aspect ratio is sent. |
| model | string | – | Image model alias: 'stable-ultra' (default, best quality), 'sd3.5', 'stable-core' (fastest/cheapest), or a full model ID |
| negative_prompt | string | – | What to exclude from the image (e.g. 'blurry, low quality') |
| output_dir | string | – | Directory to save the image. Absolute or relative to cwd. Auto-created if missing. Overrides BEDROCK_MCP_OUTPUT_DIR and default resolution. |
| prompt | string | yes | Text description of the image to generate |
| seed | integer | – | Random seed for reproducible results |
| width | number | – | Desired width (default 1024). Stability models take an aspect ratio, not pixel dimensions: width and height only pick the closest supported ratio (1:1, 16:9, 9:16, 3:2, 2:3, 4:5, 5:4, 21:9, 9:21) and… |
No output schema declared.
No examples provided.
bedrock_generate_video ~285
Start a video generation job using Luma Ray 2 on AWS Bedrock. Videos are generated asynchronously and saved to an S3 bucket. Returns a job ARN - use bedrock_video_status to check progress. Note: the output S3 bucket must be in us-west-2 (the model region). Pass start_image_path and/or end_image_path to turn local images into the first and last frame (image-to-video); the images are checked before the job is started.
| Name | Type | Req | Description |
|---|---|---|---|
| duration_seconds | – | – | Video duration in seconds: 5 (default) or 9. Luma Ray 2 supports no other value. |
| end_image_path | string | – | Local JPEG or PNG used as the last frame (512x512 to 4096x4096 pixels), absolute or relative to cwd |
| model | string | – | Video model alias: 'luma-ray' / 'ray2' (default), or a full model ID |
| prompt | string | yes | Text description of the video to generate |
| s3_uri | string | yes | S3 URI for output (e.g. 's3://my-bucket/videos/'). Bucket must be in the model's region. |
| start_image_path | string | – | Local JPEG or PNG used as the first frame (512x512 to 4096x4096 pixels), absolute or relative to cwd |
No output schema declared.
No examples provided.
bedrock_judge ~702
Send the same prompt to 2-5 Bedrock models, then have a judge model (default Claude Opus) score the answers blind - shuffled and labeled A/B/C with model names hidden - and pick a winner. Set synthesize to also get one merged answer that keeps the best of each. Pass level instead of models for three default candidates: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT-6.2 Astra, Grok 4.7); models wins when both are given, and level never changes the judge. files: Optional local text files to include in the prompt, read only from under the root folder (env BEDROCK_MCP_FILES_ROOT, default the server's working folder unless that is a home folder or a drive root): up to 10 files, 256 KB each and 1 MB in total, secret-looking files such as .env and keys and network paths are refused, and the content is sent to the model provider. A refused file fails the call before any model is billed. The files reach the candidates and the judge. The report ends with an estimated cost from list prices for the answers plus the judge call.
| Name | Type | Req | Description |
|---|---|---|---|
| criteria | string | – | What the judge scores on (default 'correctness, completeness, clarity and concision') |
| effort | string | – | Thinking depth and token spend for models that take it (Claude except Haiku 4.5, GPT-6.x, Grok 4.7, Kimi K3, Gemma 4 up to xhigh): low, medium, high, xhigh, max. Claude Opus 5.5 and Haiku 5.5 default… |
| files | array | – | Local text file paths, relative to the files root (BEDROCK_MCP_FILES_ROOT or the server's working folder). Each file is appended to the prompt and its content is sent to the model provider. |
| judge | string | – | Judge model alias or ID (default 'claude-opus') |
| level | string | – | Default candidates when models is absent: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT-6.2 As… |
| max_tokens | integer | – | Maximum tokens per candidate response |
| models | array | – | List of 2-5 candidate model aliases or IDs (e.g. ['kimi', 'nova-pro', 'deepseek']). Required unless level is given. |
| prompt | string | yes | The prompt every candidate answers |
| synthesize | boolean | – | Also return one merged final answer (default false) |
| system | string | – | Optional system prompt sent to the candidates |
| temperature | number | – | Sampling temperature for the candidates |
No output schema declared.
No examples provided.
bedrock_list_models ~76
List available Bedrock text models with their aliases, providers, use cases, and capabilities. Filter by provider name if needed.
| Name | Type | Req | Description |
|---|---|---|---|
| provider | string | – | Filter by provider (e.g. 'Anthropic', 'OpenAI', 'xAI', 'Mistral', 'Amazon', 'Qwen', 'DeepSeek') |
No output schema declared.
No examples provided.
bedrock_panel ~824
Run a short discussion between 2-5 Bedrock models and return the transcript. Unlike bedrock_compare, the models answer in turn and each one sees what the earlier ones said, so they can agree, push back and build on each other. Give each model a role (critic, builder, fact-checker, devils-advocate, simplifier, moderator, or your own words), run 1-3 rounds, and set summarize to a model that closes with where the panel agrees and disagrees. Pass level instead of models for a default panel of three: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT-6.2 Astra, Grok 4.7). Turns run one after another and each is billed: models x rounds generations, plus one for the summary. A model that fails loses its turn and the panel carries on. files: Optional local text files to include in the prompt, read only from under the root folder (env BEDROCK_MCP_FILES_ROOT, default the server's working folder unless that is a home folder or a drive root): up to 10 files, 256 KB each and 1 MB in total, secret-looking files such as .env and keys and network paths are refused, and the content is sent to the model provider. A refused file fails the call before any model is billed. The files are part of the question every turn sees. The totals line gives an estimated cost from list prices for the turns plus the summary.
| Name | Type | Req | Description |
|---|---|---|---|
| effort | string | – | Thinking depth and token spend for models that take it (Claude except Haiku 4.5, GPT-6.x, Grok 4.7, Kimi K3, Gemma 4 up to xhigh): low, medium, high, xhigh, max. Claude Opus 5.5 and Haiku 5.5 default… |
| files | array | – | Local text file paths, relative to the files root (BEDROCK_MCP_FILES_ROOT or the server's working folder). Each file is appended to the prompt and its content is sent to the model provider. |
| level | string | – | Default panel when models is absent: cheap (Nova Micro, Ministral 3 14B, Qwen3 Next 80B A3B); medium (Claude Haiku 5.5, GPT-6 Luna, Mistral Large 3 (675B)); intense (Claude Fable 5.2, GPT-6.2 Astra,… |
| max_tokens | integer | – | Maximum tokens per turn |
| models | array | – | 2-5 model aliases or IDs, in speaking order (e.g. ['nova-pro', 'kimi', 'deepseek']). Required unless level is given. A model may be listed twice only with different roles. |
| prompt | string | yes | The question or task the panel discusses |
| roles | array | – | One entry per model, in the same order: a preset (critic, builder, fact-checker, devils-advocate, simplifier, moderator), free text up to 200 characters, or null for no role |
| rounds | integer | – | Rounds of discussion, 1-3 (default 1). In round 2 and 3 each model answers again having seen everything so far. |
| summarize | string | – | Model alias or ID that closes with a short summary of where the panel agrees and disagrees. It is told it is the moderator and is not a panelist unless it is also in models. |
No output schema declared.
No examples provided.
bedrock_video_status ~59
Check the status of a Bedrock video generation job (Luma Ray 2). Returns status (InProgress, Completed, Failed) and output location.
| Name | Type | Req | Description |
|---|---|---|---|
| invocation_arn | string | yes | The job ARN returned by bedrock_generate_video |
No output schema declared.
No examples provided.
How do I install the Bedrock Multi-Model MCP server?
Bedrock Multi-Model runs locally as an npm package, launched with npx -y bedrock-multi-model-mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · bedrock-multi-model-mcp
claude mcp add sagargupta16-bedrock-multi-model-mcp -- npx -y bedrock-multi-model-mcp
What is the Bedrock Multi-Model MCP server?
Bedrock Multi-Model is an MCP server listed in the public MCP registry as io.github.Sagargupta16/bedrock-multi-model-mcp. Ask, compare and judge 35 Bedrock text models, generate and edit images, plus video and embeddings. This page covers its npm package (bedrock-multi-model-mcp).
Is the Bedrock Multi-Model MCP server safe to use?
Bedrock Multi-Model scores 89 out of 100 on VerifyMCP. It declares no install or post-install scripts. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the Bedrock Multi-Model MCP server expose?
Bedrock Multi-Model exposes 11 tools: bedrock_ask, bedrock_compare, bedrock_judge, bedrock_panel, bedrock_list_models, and 6 more. Their descriptions and schemas cost roughly 4,906 tokens of context every time the server is loaded.
Is the Bedrock Multi-Model MCP server still maintained?
Bedrock Multi-Model is still listed as active in the MCP registry. We last reached this channel on 22 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the Bedrock Multi-Model MCP server under?
Bedrock Multi-Model declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.
What did VerifyMCP test on the Bedrock Multi-Model MCP server?
VerifyMCP ran 16 checks on Bedrock Multi-Model, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.