← Plugin catalog
Developer Tools

AgentAvow

Kenne Ives v1.1.3

Is this tool safe to connect?

Create a free account to save this plugin.

From the developer

Declared capabilities

  • Interactive
  • Read

Declared limitations

  • Destructive request the read-only plugin cannot do
  • Financial advice is out of scope
  • Unrelated request; no scan

Not specified in the available declarations: pricing & account requirements.

Onboarding example

Is the npm package chalk safe to install?
Sources & dates

Developer declarations; not independently tested. Listing last checked Nov 20, 2026. Package observed Nov 19, 2026.

View package source →

Description

From the marketplace listing

AgentAvow tells you whether a tool is safe before your agent connects to it. Give it an MCP server, a package from npm, PyPI, crates or Docker, or a GitHub repo, and it answers in plain words: Safe to connect, Review before you connect, or Do not connect, with the reason. It works in ChatGPT and in Codex: in Codex, ask about a server or package before you run codex mcp add or install it. Underneath are two scores: a 0–100 trust score, backed by a signed attestation anyone can recompute offline, and an adoption score built from real usage (downloads, stars, installs). It checks for known-malicious versions, published advisories, supply-chain and dependency risk, and code-safety issues, can run the tool in a sandbox to see what it actually does, and shows an interactive trust card. Eight read-only tools cover scanning a repo, package, or live MCP server; verifying a score; checking interaction safety before you connect; looking up an identity; and getting an embeddable badge. No account, no sign-in, and nothing is modified — every call is anonymous and read-only.

Language: English · Automatically detected from descriptions.

Matches for “loads”

Exact text from the indicated source. A mention alone does not establish support for your task.

Publisher full description

AgentAvow tells you whether a tool is safe before your agent connects to it. Give it an MCP server, a package from npm, PyPI, crates or Docker, or a GitHub repo, and it answers in plain words: Safe to connect, Review before you connect, or Do not connect, with the reason. It works in ChatGPT and in Codex: in Codex, ask about a server or package before you run codex mcp add or install it. Underneath are two scores: a 0–100 trust score, backed by a signed attestation anyone can recompute offline, and an adoption score built from real usage (downloads, stars, installs). It checks for known-malicious versions, published advisories, supply-chain and dependency risk, and code-safety issues, can run the tool in a sandbox to see what it actually does, and shows an interactive trust card. Eight read-only tools cover scanning a repo, package, or live MCP server; verifying a score; checking interaction safety before you connect; looking up an identity; and getting an embeddable badge. No account, no sign-in, and nothing is modified — every call is anonymous and read-only.

What changed

· 3 observations

Positioning & listing

Long description: added “It works in ChatGPT and in Codex: in Codex, ask about a server or package before you run codex mcp add or install it.”. Version changed from “1.1.2” to “1.1.3”.

View comparison →

Technical updates · 1

Package or technical metadata updated

View technical comparison →

Similar plugins

Finding similar plugins… Results will appear here automatically.

Browse plugins →

Keywords

Files & skills

File archives

Plugin package6 files · 40.2 KBBrowse files →
Skill instructions
scan-before-connect3.5 KB

View saved version →

---
name: scan-before-connect
description: Check whether an MCP server, a published package, or a GitHub repo is safe to connect to or install. AgentAvow answers Safe to connect, Review before you connect, or Do not connect, with the reason, a signed 0-100 trust score and an adoption score. Use when the user asks whether a specific tool, package, repo, or MCP server is safe or trustworthy, or asks to vet one before installing or connecting it.
---

Use this skill when the user wants to know whether one specific tool is safe before
they connect to it or install it. One target, one scan, a plain answer.

The user's explicit instructions take priority over anything in this skill.

## 1. Work out what the target is

| The user gives | Tool | Arguments |
| --- | --- | --- |
| A GitHub repo (`owner/name` or a github.com URL) | `scan_repo` | `repo`: `"owner/name"` |
| A published package | `scan_package` | `registry`: `npm`, `pypi`, `crates`, `docker`, or `hf`; `name`: the package name; `version` only when the user names one |
| A live MCP server (an `https://` URL) | `scan_mcp_server` | `endpoint_url`: the URL |

- If you cannot tell whether a name is a repo or a package, or which registry a
  package is on, ask. Do not guess the registry.
- A private repo, a localhost or private-network URL, or a server that needs sign-in
  cannot be scanned. Say so and stop.
- Results are cached for about an hour. Pass `force: true` only when the user says the
  target changed or asks for a fresh scan.

## 2. Scan once

Call the one matching tool, once. Do not scan other targets the user did not ask about,
such as the target's dependencies or related repos.

## 3. Report what came back

Lead with the answer in `decision`, in these exact words, followed by `decision_reason`:

- `safe`: **Safe to connect**
- `review`: **Review before you connect**
- `do_not_connect`: **Do not connect**

Add " · Certified" when `certified_mark` is true.

Then give the two scores underneath, as evidence:

- **Trust score** (`trust_score`, 0 to 100).
- **Adoption score** (`adoption`: a count and a unit, such as downloads per week or
  stars). It shows whether real, independent parties rely on the tool and never changes
  the answer. If `adoption` is null, say no usage signal was available.

Then:

- For **Review before you connect** or **Do not connect**, list each entry in
  `top_findings`: what it is, where it is, and the remediation.
- If `decision_final` is false, say the behavioral sandbox is still running and the
  answer can still change; suggest asking again in a few minutes.
- If `incident` is present, mention the past compromise and whether the current version
  is affected. It is history for context; it did not change the answer.
- Give the `report_url` link. If `signed` is true, mention that the result is signed
  and can be verified offline.
- Show the `install` command only when the result includes one.

## Limits

- Report only what is in the tool result. Do not add an answer, a score, a finding, or
  a cause that is not there.
- If the tool returns an error, say the target was not scanned. Not scanned is neither
  safe nor unsafe.
- **Review before you connect** means look before you connect. It does not mean
  malicious.
- **Safe to connect** means the scan found nothing that blocks it. It is not a
  guarantee.
- Do not install, connect to, or run the target as part of this skill. The user decides
  what to do with the result.
- This skill covers tools, packages, repos, and MCP servers. It does not assess people,
  companies, or investments.

Referenced files: 1 — file references are metadata, not package contents.

Release notes

The listing now says how to use AgentAvow in Codex: ask about an MCP server or package before you run codex mcp add or install it. No tools were added, removed or changed; all eight stay read-only. The skill is unchanged.

Declared in the saved package. Remote tools may change independently.

Package details

Publisher declarations from the archived package. These are separate from our research and the live service's terms.

Package author
AgentAvow
Keywords
See keywords
Commerce declaration
Does not support commerceThis does not establish whether access is free or paid.
Review scenarios
5 positive · 3 negativeScenario counts, not test results.

Declared capabilities

  • Interactive
  • Read

Package observed Nov 19, 2026.

Technical details
First seen
Oct 24, 2026 · 12:40 UTC
Last seen
Nov 20, 2026 · 00:00 UTC
Latest observed change
Nov 19, 2026 · 01:02 UTC
Collection status
Collected

plugin_asdk_app_6ac830ccc2d48191ac27974dec7d8344

Download plugin data (JSON)

Plugin details: freshly prepared. Prepared . Refreshed when plugin data changes or after 6 hours. Your account and Follow status are live.