Skip to content
verify mcp Beta VerifyMCP is currently in beta. If you notice any issues, get in touch and we’ll put it right.

hush

NPM · @OMAREI/HUSH

Team secrets your AI agent can use but never read: encrypted in the repo, value-blind tools.

Available components

0 this week 92 Trust /100

Last scanned (15 hours ago)

Scores higher than 92% of npm servers

Category Security

Trust breakdown (7 categories)

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 Security100
  • No malware found by supply-chain analysis.Pass
  • No known CVEs affecting this package version or its production dependencies.Pass
  • No install/post-install scripts declared.Pass
  • No production dependencies, so there is no dependency health to assess. View diagnostics → Pass
Provenance & Transparency100
  • Source repository is publicly reachable at the declared URL. View diagnostics → Pass
  • Cryptographically verified build provenance (signed, bound to omarei-omoto/hush). View diagnostics → Pass
  • Clear OSI-approved license (MIT).Pass
  • Actively maintained (last published 33 days ago).Pass
  • Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability67
  • AI-judged instruction clarity (excellent).Pass
  • Context-footprint check failed: tool/resource definitions use about 1965 tokens (~178/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 Management100
  • No destabilizing schema changes in the last 30 days.Pass
Tool Coverage96
  • 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
  • 89% of tool parameters carry a description.Partial
Tool Safety75
  • No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
  • 0 of 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation; "hush_push" implies "send" and declares no destructiveHint at all, which the MCP spec reads as destructive by default. See how to fix → Fail
  • An AI judge read all 12 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
What we tested

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 →

Changelog

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.

  • 14 Nov 26 0
    • Stability: 0.97 → pass security
  • 13 Nov 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 93 to 97. That category is still filling its 30-day observation window: 28 days of observed history at the previous scan, 29 at this one. The score rises as the window fills, whether or not the server changes.

  • 11 Nov 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 87 to 90. That category is still filling its 30-day observation window: 26 days of observed history at the previous scan, 27 at this one. The score rises as the window fills, whether or not the server changes.

  • 10 Nov 26 0
    • Package version: 0.13.1 → 0.13.2 functional
  • 9 Nov 26 +3
    • Source repository: unverified → pass ▲ security
    • Stability: pass → 0.83 functional
  • 8 Nov 26 −5
    • Provenance: pass → unverified ▼ security
  • 30 Oct 26 0
    • Stability: 0.97 → pass security
  • 29 Oct 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 93 to 97. That category is still filling its 30-day observation window: 28 days of observed history at the previous scan, 29 at this one. The score rises as the window fills, whether or not the server changes.

Diagnostics

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 21 Nov 2026 · Analysed npm/@omarei/hush@0.13.2

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 omarei-omoto/hush
Certificate issuer https://token.actions.githubusercontent.com
Certificate SAN https://github.com/omarei-omoto/hush/.github/workflows/release.yml@refs/tags/v0.13.2
Rekor log index 3310441986
Predicate type SLSA build provenance https://slsa.dev/provenance/v1
Subject digest sha512:7e09312cdb238395441c6bb9541c694f46e86cbb7f4a514e107ae256d01769b07ab7c4ecfe63206d37754d0acee7b90f860906f75254887741fe54934

Background: How many MCP packages publish verified provenance →

Dependencies 0 packages
Packages resolved 0
Tree resolution Complete

Background: SBOMs and build attestations, explained →

MCP tools · 11 exposed · ~1,885 tokens

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 →

Tool Tokens
hush_add_secret ~278

Add a credential to the vault WITHOUT it passing through this conversation. A secure input box opens on the user's screen; they paste the value there and it is encrypted straight into the vault. You get back only a confirmation. ALWAYS use this instead of asking the user to paste a key into the chat. Give a set name (created if it doesn't already exist) and either a known service (hush knows which variables it needs, e.g. 'fal') or a bare key.

NameTypeReqDescription
accountstring–Deprecated (removed in hush 2.0), used with service — together they mean set "service/account".
envstring–Deprecated alias for "set", removed in hush 2.0.
keystring–A single variable name, if this isn't a known service.
servicestring–Deprecated alias (removed in hush 2.0): e.g. 'fal'. hush knows which variables it needs.
setstring–The set to add this to, e.g. 'work-fal'. Created if absent. Default: the project's default set.
wherestring–Where to create a new set: this project's vault (default) or the user's own library.
whystring–Shown to the user so they know what they are approving.

No output schema declared.

No examples provided.

hush_check_repo ~73

Scan the current codebase for the environment variables it references, and report which are present in the vault and which are missing. Use this before running or building a project to find out whether its configuration is complete.

NameTypeReqDescription
envstring––
pathstring–Directory to scan (default: project root).

No output schema declared.

No examples provided.

hush_describe_secret ~78

Describe one secret without revealing it: whether it exists, its length, a masked preview, who last set it and when. Use this to confirm a credential is configured.

NameTypeReqDescription
envstring–Deprecated alias for "set", removed in hush 2.0.
keystringyes–
setstring––

No output schema declared.

No examples provided.

hush_list_accounts ~46

Deprecated (removed in hush 2.0), use hush_list_sets — this returns exactly the same thing. Kept registered so a skill file written before sets replaced accounts still works.

Input schema present but exposes no named parameters.

No output schema declared.

No examples provided.

hush_list_secrets ~95

List the NAMES of secrets available in one named set of the current project's vault. Returns names, which set they live in, and when they were last changed. Never returns secret values — use hush_run to actually use a secret.

NameTypeReqDescription
envstring–Deprecated alias for "set", removed in hush 2.0.
setstring–Which set to list (default: the project's default set).

No output schema declared.

No examples provided.

hush_list_sets ~113

List every named set the agent could use with this project: sets in the user's own library (global, e.g. 'Personal fal', 'Work fal') and sets in the project's own vault. Each entry says where it lives (library or project), whether this project already uses it, and its key names. Call this when the user names a set loosely ("use my acme fal key", "the work fal account") so you can pass the right name to hush_run. Returns names only, never values.

Input schema present but exposes no named parameters.

No output schema declared.

No examples provided.

hush_provision ~165

Prepare a CLI or codebase to run with the right credentials. Give it a tool name (e.g. 'wrangler') and optionally which set to use. It works out which service the tool authenticates with, checks which of this project's used sets already provide it, prompts the user to add it if missing, and tells you the exact hush_run call to make. Use this when the user says something like 'set up wrangler with my personal Cloudflare set'.

NameTypeReqDescription
servicestring–Override the detected service.
setstring–Which set to check, e.g. 'work-fal'. Default: whichever of this project's used sets already provide it.
toolstringyesThe command that needs credentials, e.g. 'wrangler'.

No output schema declared.

No examples provided.

hush_push ~297

Send one of the project's secrets to where the app runs — Cloudflare Workers or Pages, Vercel, Convex, Fly.io, Railway, Firebase Functions, GitHub Actions, GitLab CI, Google Cloud Secret Manager, Koyeb, Fastly, Pulumi or SST — through that host's own CLI, with the value on its standard input. The value never passes through you or a command line, and hush refuses a value the host's CLI would change. The user approves each push. Use this instead of telling the user to paste a key into a dashboard, and instead of `sh -c '... | wrangler secret put'`.

NameTypeReqDescription
asstring–A different name for it there. Default: the key's own name.
deployboolean–Fly and Railway only: deploy as part of setting it. Default: stage it for the next deploy.
envstring–The host's environment, stage or deployment, e.g. 'production' for Vercel.
keystringyesThe key to send, e.g. 'STRIPE_SECRET_KEY'.
targetstring–Which Worker, Pages project, app, repository, project, stack or store. Omit to use the one the project's own config names (wrangler.jsonc, fly.toml, .firebaserc), where it has one.
tostringyesThe host.

No output schema declared.

No examples provided.

hush_request ~312

Make an authenticated HTTP request without the credential ever entering this conversation. Write $NAME where a secret belongs and hush substitutes it from the vault inside its own process; you get back the response with any secret values masked as [redacted:NAME]. Use this for any API call that needs a key, instead of trying to run curl/wget (denied) or building an Authorization header yourself. Secrets go into HEADER VALUES only unless you also list 'body' or 'query' in substitute. https only; the host must be one the policy allows.

NameTypeReqDescription
bodystring–Request body. Substituted only if "body" is in substitute.
headersobject–Header name to value. Put the secret placeholder in the value, e.g. {"Authorization": "Bearer $STRIPE_KEY"}. The value is substituted inside hush and never returned to you.
methodstring–HTTP method. Default GET, or POST when a body is given.
setsarray–Extra named sets to layer on top of this project's usual ones, for this call only, e.g. ["work-stripe"]. Later entries win on a shared key name. Use hush_list_sets to see the options.
substitutearray–Extra places, besides header values, where $NAME may be replaced. Omit for header values only, which is the safe default.
urlstringyesFull URL, e.g. "https://api.stripe.com/v1/refunds".

No output schema declared.

No examples provided.

hush_run ~308

Run a shell command with the project's secrets injected as environment variables. The command sees real credentials; you only see its output, with any secret values masked. This is the correct way to perform a task that needs a credential (migrations, deploys, authenticated API calls) without the credential entering this conversation.

NameTypeReqDescription
accountsobject–Deprecated alias for sets (removed in hush 2.0): {"fal":"acme"} means the set named "fal/acme". Prefer sets.
argsarray–Arguments, e.g. ['run','build'].
commandstringyesExecutable to run, e.g. 'npm'.
cwdstring––
envstring–Deprecated (removed in hush 2.0): the base set for this run, layered under sets. Prefer sets.
setsarray–Extra named sets to layer on top of this project's usual ones, for this run only, e.g. ["work-fal"]. Later entries win over earlier ones on a shared key name. Use hush_list_sets to see the options. O…
stdinstring–The name of one key whose value is written to the command's standard input — for a CLI that reads a secret from stdin (e.g. `wrangler secret put NAME`). No shell needed, and the value is never on a c…

No output schema declared.

No examples provided.

hush_setup_status ~120

What is set up for hush on this machine and in this project, and what is next. Each step has a status, the exact command that does it, and a kind: auto (run it), choice (ask the user which option, then run it), or person (run it and wait: it asks the user on their screen, and you must never answer it or add --yes). Call this when the user asks you to set up hush, or when another hush tool says hush is not set up here. It never shows a secret.

Input schema present but exposes no named parameters.

No output schema declared.

No examples provided.

Install

How do I install the hush MCP server?

hush runs locally as an npm package, launched with npx -y @omarei/hush 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 · @omarei/hush

# add to Claude Code
claude mcp add omarei-omoto-hush -- npx -y @omarei/hush mcp
Common questions

What is the hush MCP server?

hush is an MCP server listed in the public MCP registry as io.github.omarei-omoto/hush. Team secrets your AI agent can use but never read: encrypted in the repo, value-blind tools. This page covers its npm package (@omarei/hush).

Is the hush MCP server safe to use?

hush scores 92 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 21 November 2026. 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 hush MCP server expose?

hush exposes 11 tools: hush_list_secrets, hush_list_sets, hush_list_accounts, hush_setup_status, hush_check_repo, and 6 more. Their descriptions and schemas cost roughly 1,885 tokens of context every time the server is loaded.

Is the hush MCP server still maintained?

hush is still listed as active in the MCP registry. We last reached this channel on 21 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 hush MCP server under?

hush 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 hush MCP server?

VerifyMCP ran 16 checks on hush, 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.