# io.github.tponscr-debug/oracle-h (npm · oracle-h-mcp)

Human approval gate for AI agents. Critical actions approved or rejected by a human via Telegram.

- Trust score: 52/100 (low)
- Change this week: +15
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- npm · `oracle-h-mcp`: 52/100 (this document), [markdown](https://verifymcp.io/servers/tponscr-debug-oracle-h/oracle-h-mcp.md), [page](https://verifymcp.io/servers/tponscr-debug-oracle-h/oracle-h-mcp)

## Channel facts

- Registry: `npm`
- Package: `oracle-h-mcp`
- Version: `1.0.0`
- 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**: 86/100
  - No malware found by supply-chain analysis.
  - Only part of the dependency tree could be resolved (94 of 98), 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 (94 of 98), so this covers what we could see, not the whole tree.
- **Provenance & Transparency**: 19/100
  - Repository check failed: the declared repository URL redirects; it must resolve directly.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 148 days ago).
  - Security-disclosure policy not yet verified: we couldn't inspect the source repository.
- **Schema Quality & AI Usability**: 22/100
  - AI-judged instruction clarity (poor).
  - Context-footprint check failed: tool/resource definitions use about 630 tokens (~315/item across 2 items; 2 tools + 0 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**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add tponscr-debug-oracle-h -- npx -y oracle-h-mcp
```

### Codex

```bash
codex mcp add tponscr-debug-oracle-h -- npx -y oracle-h-mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "tponscr-debug-oracle-h": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "oracle-h-mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add tponscr-debug-oracle-h --command npx --arg -y --arg oracle-h-mcp
```

### Hermes

```yaml
mcp_servers:
  tponscr-debug-oracle-h:
    command: "npx"
    args: ["-y", "oracle-h-mcp"]
```

### Other

```json
{
  "mcpServers": {
    "tponscr-debug-oracle-h": {
      "command": "npx",
      "args": [
        "-y",
        "oracle-h-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.

### 2026-08-03 (score 52, +3)

- [functional improvement] Stability: unverified → 0.23

### 2026-08-02 (score 49, +13)

- [security regression] Known CVEs: partial → unverified
- [security improvement] Malware scan: unverified → pass
- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Dependency health: partial → unverified
- [functional improvement] Schema quality: unverified → poor

### 2026-08-01 (score 36, +36)

- [security regression] Provenance: unverified → fail
- [security improvement] Install scripts: unverified → pass
- [security improvement] Known CVEs: unverified → partial
- [security] Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window).
- [functional improvement] MCP protocol: unverified → pass
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] License: unverified → pass
- [functional improvement] Dependency health: unverified → partial
- [functional improvement] Maintenance: unverified → pass
- [functional] Licence: MIT

### 2026-07-31 (score 0, −18)

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

### 2026-07-30 (score 18, −19)

- [functional regression] Tool coverage: 100 → unverified
- [functional] First check of Schema quality: unverified

### 2026-07-27 (score 37)

First indexed and scored.

## MCP tools (2)

### `oracle_validate` (~433 tokens)

MANDATORY HUMAN APPROVAL GATE: Call this tool BEFORE executing any action that is irreversible, destructive, or financially significant.

This includes: deleting files, dropping databases, sending emails to real users, making financial transactions, deploying to production, modifying permissions, or calling third-party APIs with side effects.

USE THIS TOOL WHEN:
\- Action cannot be undone (file deletion, DB drop, git force-push, data wipe)
\- Action has financial implications (payments, purchases, billing changes)
\- Action affects real users or external systems (emails, API write calls)
\- Action modifies infrastructure (server config, DNS, deployments, permissions)
\- Action scope exceeds what was explicitly authorized in the current session

DO NOT USE FOR: Read-only operations, local sandboxed operations, actions explicitly pre-approved in system prompt.

Returns a request_id. Call oracle_poll_status every 10-15 seconds until "approved" or "rejected". DO NOT proceed until terminal status. If unreachable, treat as BLOCKED.

Input parameters:

- `action` (string, required): Precise technical description of the exact action about to be executed. Include the target (file path, database name, API endpoint), the operation type, and scope. Example: 'DROP TABLE users in produ…
- `expires_in_seconds` (number): Seconds to wait for human response. Use 60-120 for urgent real-time workflows. Use 300-600 for async tasks. If expired without decision, treat as rejected and abort.
- `financial_amount` (number): Total financial amount at stake in USD. Set to 0 if no financial impact. Include transaction amount or potential cost of mistakes.
- `projected_impact` (string, required): Concrete description of what goes wrong if this action fails or is incorrect. Include: data loss risk, affected users/systems, recovery difficulty. Example: 'Permanent deletion of 50,000 user records…
- `rationale` (string, required): Why this action is necessary to accomplish the goal, and why this specific approach was chosen over safer alternatives. Be honest about uncertainty.

### `oracle_poll_status` (~197 tokens)

POLLING PROTOCOL (mandatory): Check the human decision for a request submitted via oracle_validate.

1\. Call this with the request_id from oracle_validate
2\. If status is "pending": wait 10-15 seconds, then call again
3\. Repeat until status is "approved" OR "rejected" OR "expired"
4\. NEVER proceed with the action while status is "pending"

DECISION RULES (non-negotiable):
\- "approved" → Human confirmed. You may now execute the action.
\- "rejected" → Human denied. DO NOT execute. Inform the user and stop.
\- "expired"  → No response in time. Treat as rejected. DO NOT execute.
\- "pending"  → Still waiting. Poll again in 10-15 seconds.

Input parameters:

- `request_id` (string, required): The unique request_id string returned by oracle_validate. Copy it exactly from the oracle_validate response.

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/tponscr-debug-oracle-h/oracle-h-mcp#diagnostics

## Score history

- 2026-08-03: 52
- 2026-08-02: 49
- 2026-08-01: 36
- 2026-07-31: 0
- 2026-07-30: 18
- 2026-07-28: 37
- 2026-07-27: 37

## Links

- npm package: https://www.npmjs.com/package/oracle-h-mcp
- Socket report: https://socket.dev/npm/package/oracle-h-mcp
- Changelog RSS feed: https://verifymcp.io/servers/tponscr-debug-oracle-h/oracle-h-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/tponscr-debug-oracle-h/oracle-h-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/tponscr-debug-oracle-h/oracle-h-mcp