# io.github.commet/argus-decision-mcp (npm · argus-decision-mcp)

Preserve a user-authorized judgment and append later answers. No model verdict or human score.

- Trust score: 61/100 (medium)
- Change this week: +53
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- npm · `argus-decision-mcp`: 61/100 (this document), [markdown](https://verifymcp.io/servers/commet-argus-decision-mcp/argus-decision-mcp.md), [page](https://verifymcp.io/servers/commet-argus-decision-mcp/argus-decision-mcp)

## Channel facts

- Registry: `npm`
- Package: `argus-decision-mcp`
- Version: `2.0.20`
- Transport: `stdio`

## Trust breakdown

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. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Supply Chain Security**: 87/100
  - No malware found by supply-chain analysis.
  - Only part of the dependency tree could be resolved (97 of 101), so this covers what we could see, not the whole tree.
  - No install/post-install scripts declared.
  - Only part of the dependency tree could be resolved (97 of 101), so this covers what we could see, not the whole tree.
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Build provenance is cryptographically sound, but it attests a different repository to the one declared in the registry. Most often the declared URL is simply stale.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 3 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 49/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (poor).
  - Context-footprint check failed: tool/resource definitions use about 2260 tokens (~282/item across 8 items; 6 tools + 2 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 23/100
  - Stability observed for 7 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 81/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 33% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add commet-argus-decision-mcp -- npx -y argus-decision-mcp
```

### Codex

```bash
codex mcp add commet-argus-decision-mcp -- npx -y argus-decision-mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "commet-argus-decision-mcp": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "argus-decision-mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add commet-argus-decision-mcp --command npx --arg -y --arg argus-decision-mcp
```

### Hermes

```yaml
mcp_servers:
  commet-argus-decision-mcp:
    command: "npx"
    args: ["-y", "argus-decision-mcp"]
```

### Other

```json
{
  "mcpServers": {
    "commet-argus-decision-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "argus-decision-mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-02 (score 61, +39)

- [security improvement] Install scripts: unverified → pass
- [security improvement] Known CVEs: unverified → partial
- [security improvement] Malware scan: unverified → pass
- [security] Provenance: Build provenance is cryptographically sound, but it attests a different repository to the one declared in the registry. Most often the declared URL is simply stale.
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Schema quality: 100 → unverified
- [functional improvement] Schema quality: unverified → poor
- [functional improvement] MCP protocol: unverified → pass
- [functional improvement] Maintenance: unverified → pass
- [functional improvement] Stability: unverified → 0.20
- [functional improvement] License: unverified → pass
- [functional improvement] Dependency health: unverified → partial
- [functional] Licence: MIT

### 2026-08-01 (score 22, −7)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-31 (score 29, 0)

- [functional] Package version: 2.0.19 → 2.0.20

### 2026-07-30 (score 29, +23)

- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] Schema quality: unverified → 100
- [functional] First check of Schema quality: fail
- [functional] First check of Tool coverage: 33
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: unverified
- [functional] First check of Tool coverage: 100
- [functional] Package version: 2.0.13 → 2.0.19
- [functional] Package version: 2.0.13 → 2.0.17

### 2026-07-29 (score 6, 0)

- [security improvement] Known CVEs: unverified → partial
- [security improvement] Install scripts: unverified → pass
- [security] Provenance: Build provenance is cryptographically sound, but it attests a different repository to the one declared in the registry. Most often the declared URL is simply stale.
- [functional improvement] License: unverified → pass
- [functional improvement] Maintenance: unverified → pass
- [functional improvement] Dependency health: unverified → partial
- [functional] Licence: MIT
- [functional] Package version: 2.0.8 → 2.0.13
- [functional] Package version: 2.0.8 → 2.0.12
- [functional] Package version: 2.0.8 → 2.0.10
- [functional] Package version: 2.0.8 → 2.0.9

### 2026-07-28 (score 6, −2)

- [functional regression] Security disclosure: unverified → fail
- [functional regression] Dependency health: partial → unverified
- [functional] Package version: 2.0.1 → 2.0.8
- [functional] Package version: 2.0.1 → 2.0.2

### 2026-07-27 (score 8)

First indexed and scored.

## MCP tools (6)

### `argus_capture` (~490 tokens)

결정 다루기 · Work with a decision

결정을 대신 내리지 않고, 그 결정이 어떤 전제 위에 서 있는지와 아직 열린 질문을 사용자의 말 그대로 포착합니다. 기록된 결정의 전제·현재 사실·예측 문장·상태도 관리합니다.

Capture a decision and its user-stated context without deciding for the user. Preserve at most one load-bearing assumption; tag an AI-drafted premise as ai_surfaced.

Input parameters:

- `action` (string, required): 수행할 세부 작업입니다. The operation to perform.
- `already_decided` (boolean)
- `amendment` (string): amend_context에서 전제를 고치는 방식입니다. How to amend a premise for action=amend_context.
- `apply_to_matching` (boolean)
- `argus_dir` (string)
- `changed` (boolean)
- `check_by` (string): 예측을 현실과 대조할 미래 확인일(YYYY-MM-DD)입니다. Future date, in YYYY-MM-DD, when the prediction can be checked.
- `decision` (string)
- `dismiss_reason` (string)
- `external` (boolean)
- `finding` (string)
- `id` (string): 결정 식별자입니다. Stable identifier for the decision.
- `load_bearing` (boolean)
- `load_bearing_assumption` (string)
- `monitoring_enabled` (boolean)
- `note` (string)
- `numeric_value` (number)
- `predicate` (string): 현실이 참/거짓으로 확인할 수 있는 예측입니다. A falsifiable prediction that reality can answer.
- `premises` (array): 추가할 전제 또는 미결 질문 목록입니다. Premises or open questions to add.
- `recheck_cadence_days` (integer)
- `reconsider_cadence_days` (integer)
- `ref` (string)
- `related_to` (array)
- `reversibility` (string)
- `source` (string): action=update_fact에서만 사용: 현재 사실을 확인한 출처(url | user_stated | host_reported). For update_fact only.
- `source_detail` (string)
- `stakes` (string)
- `status_quo` (string)
- `text` (string): 사용자가 고친 전제 원문입니다. The user’s corrected premise wording.

Output parameters:

- `data` (object)
- `error_code` (string)
- `invalid_fields` (array)
- `message` (string)
- `next_actions` (array)
- `ok` (boolean)
- `over_fire_gate` (object)
- `recovery` (string)
- `surface` (string)
- `tool` (string)

### `argus_predict` (~480 tokens)

예측 저장 · Seal a prediction

현실이 확인할 수 있는 예측과 확인일을 저장하고 작성 주체를 정직하게 기록합니다.

Record one falsifiable prediction and when reality can answer it. Use the user’s words and offer confirmation once.

Input parameters:

- `argus_dir` (string)
- `basis` (string)
- `check_by` (string, required): 예측을 현실과 대조할 미래 확인일(YYYY-MM-DD)입니다. Future date, in YYYY-MM-DD, when the prediction can be checked.
- `confirm_draft` (boolean): AI가 만든 예측 초안을 사용자에게 한 번에 확인받을 때 사용합니다. Ask for one-tap confirmation when the AI drafted the prediction. Optional extra confirmation: force the one-tap confirm even for a "user" predicate. ai_surfac…
- `human_judgment` (string)
- `human_only` (string)
- `id` (string, required): 결정 식별자입니다. Stable identifier for the decision. A short slug you pick for this decision (e.g. "q3-cutover"). A fresh id starts the record on its own.
- `predicate` (string, required): 현실이 참/거짓으로 확인할 수 있는 예측입니다. A falsifiable prediction that reality can answer. A prediction reality can mark true/false. Good: "cutover downtime < 5 min". Bad: "it will go well".
- `predicate_owner` (string, required): 예측 문장의 작성 주체입니다. 출처를 꾸미지 않습니다. Who authored the prediction. Never forge provenance. Provenance. Never forge. "user" = the user wrote or affirmed it. "ai_surfaced" = Argus drafted, unconfirmed — on…
- `real_question` (string)
- `unverified_assumption` (string)

Output parameters:

- `data` (object)
- `error_code` (string)
- `invalid_fields` (array)
- `message` (string)
- `next_actions` (array)
- `ok` (boolean)
- `over_fire_gate` (object)
- `recovery` (string)
- `surface` (string)
- `tool` (string)

### `argus_check_in` (~75 tokens)

확인할 기록 보기 · Check what is due

확인일이 됐거나 다시 살펴볼 결정·전제·미결 질문을 보여줍니다.

Show only open records that need attention now. Reads open records; the first call may initialize the ledger folder.

Input parameters:

- `argus_dir` (string)
- `include_upcoming_days` (integer)

Output parameters:

- `data` (object)
- `error_code` (string)
- `invalid_fields` (array)
- `message` (string)
- `next_actions` (array)
- `ok` (boolean)
- `over_fire_gate` (object)
- `recovery` (string)
- `surface` (string)
- `tool` (string)

### `argus_resolve` (~378 tokens)

실제 결과 기록 · Settle against reality

확인일이 된 예측에 실제로 일어난 일을 기록합니다. Argus는 결과를 평가하지 않습니다.

Record an outcome the user explicitly stated for a tracked prediction. Reality answers; Argus never grades.

Input parameters:

- `argus_dir` (string)
- `broken_premise_ref` (string)
- `defer_to` (string)
- `id` (string, required): 결정 식별자입니다. Stable identifier for the decision.
- `outcome` (string): 예측에 대해 현실에서 실제로 일어난 결과입니다. 모델이 추론하지 않습니다. What actually happened to the prediction. The model must not infer it. What reality did to the prediction — record the user's words, never infer. Definitio…
- `outcome_source` (string): 결과를 말한 주체입니다. 현재는 user_stated만 허용합니다. Who stated the result. Only user_stated is accepted. Always "user_stated" — an AI-inferred outcome cannot be expressed. Defaulted, so you may omit it.
- `what_happened` (string): 실제로 일어난 일을 사용자의 말로 기록합니다. What actually happened, in the user’s own words. What reality did, in the user's words. Required when recording held/avoided/partial/missed. Omit for still_pending and pas…

Output parameters:

- `data` (object)
- `error_code` (string)
- `invalid_fields` (array)
- `message` (string)
- `next_actions` (array)
- `ok` (boolean)
- `over_fire_gate` (object)
- `recovery` (string)
- `surface` (string)
- `tool` (string)

### `argus_patterns` (~209 tokens)

판단 기록 보기 · View judgment history

진행 중인 결정, 전체 기록, 판단 영수증, 전제, 누적 기록을 읽고, 당신이 쓴 예측·전제와 그 결과를 되읽는 회고를 봅니다.

Read active decisions, receipts, timelines, and recurring patterns. Read-only; history is context, never a verdict.

Input parameters:

- `argus_dir` (string)
- `id` (string): 결정 식별자입니다. Stable identifier for the decision.
- `view` (string): 불러올 기록의 종류입니다. Which part of the decision record to read. active는 진행 중인 결정, all은 전체 기록, receipt는 판단 영수증, decision_context는 결정의 전제와 미결 질문, timeline은 누적 정산 결과 요약(예측대로·걱정 피함·일부·빗나감 빈도), reflection은 당신…

Output parameters:

- `data` (object)
- `error_code` (string)
- `invalid_fields` (array)
- `message` (string)
- `next_actions` (array)
- `ok` (boolean)
- `over_fire_gate` (object)
- `recovery` (string)
- `surface` (string)
- `tool` (string)

### `argus_settings` (~135 tokens)

Argus 설정 · Argus settings

언어와 알림, 전제 동기화 설정을 관리하고 필요할 때 계정 기록을 동기화합니다.

Read or update language, reminders, and explicit sync settings. Argus initializes the current project on first use.

Input parameters:

- `action` (string, required): 수행할 세부 작업입니다. The operation to perform.
- `ambient_mute` (boolean)
- `argus_dir` (string)
- `due_only` (boolean)
- `import_settlements` (boolean)
- `locale` (string)
- `premise_sync` (boolean)
- `push_local` (boolean)

Output parameters:

- `data` (object)
- `error_code` (string)
- `invalid_fields` (array)
- `message` (string)
- `next_actions` (array)
- `ok` (boolean)
- `over_fire_gate` (object)
- `recovery` (string)
- `surface` (string)
- `tool` (string)

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/commet-argus-decision-mcp/argus-decision-mcp#diagnostics

## Score history

- 2026-08-03: 61
- 2026-08-02: 61
- 2026-08-01: 22
- 2026-07-31: 29
- 2026-07-30: 29
- 2026-07-29: 6
- 2026-07-28: 6
- 2026-07-27: 8

## Links

- npm package: https://www.npmjs.com/package/argus-decision-mcp
- Socket report: https://socket.dev/npm/package/argus-decision-mcp
- Repository: https://github.com/commet/argus
- Website: https://argus.voyage/
- Changelog RSS feed: https://verifymcp.io/servers/commet-argus-decision-mcp/argus-decision-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/commet-argus-decision-mcp/argus-decision-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/commet-argus-decision-mcp/argus-decision-mcp