SuperAPI AI Directory

clairvoyance

Security & auditing

View repository

Updated 10-20 16:15

About this skill

Clairvoyance

Get a snapshot of a remote system or environment you cannot directly access.

What This Skill Does

In D&D, Clairvoyance lets you place an invisible sensor in a location you know, seeing or hearing through it. Unlike Scrying (which follows a specific target in real time), Clairvoyance is a point-in-time snapshot of a place. The real-world version is remote system inspection: checking the state of a production environment you cannot SSH into, reading the public-facing state of a competitor's deployment, or gathering the current observable state of a system through its exposed interfaces without modifying it. In this grimoire, Clairvoyance is treated as a hybrid spell with a shipping-now delivery profile. Canonical reference input: Clairvoyance (spell).

When To Use

  • You need to understand the current state of a system or environment you cannot directly access or modify.
  • You want a snapshot assessment — not ongoing monitoring (that is Scrying) but a point-in-time read.
  • You need to gather observable information about a remote service, API, or public-facing system.

Prerequisites

  • No extra runtime dependencies beyond Hermes Agent and the normal toolset for this session.

Procedure

  1. Restate the target, the success condition, and any no-touch boundaries before taking action.
  2. Identify the target location or system and what specifically you want to observe.
  3. Determine what is observable through legitimate interfaces: public APIs, status pages, DNS records, HTTP headers, documented endpoints.
  4. Collect the snapshot: gather all available observable state at this moment.
  5. Distinguish facts (what is directly observed) from inferences (what the observations suggest about internal state).
  6. Deliver the snapshot with a timestamp and confidence levels for any inferences.
  7. Package the result as the deliverables below, with confidence, assumptions, and unresolved risk called out explicitly.

Deliverables

  • A point-in-time snapshot of the target's observable state.
  • Inferences about internal state, clearly marked as inferences with confidence levels.
  • Recommendations for what to investigate further based on what the snapshot reveals.

Pitfalls / Guardrails

  • Keep the theatrical framing, but name the concrete mechanism that makes the skill useful right now.
  • Clairvoyance observes through legitimate channels only. Do not attempt to access systems beyond their public or authorized interfaces.
  • A snapshot is a moment in time, not a guarantee of current state. Always timestamp observations and note that they may already be stale.

Verification

  • Check that the result includes every deliverable promised above.
  • Check that confirmed facts, assumptions, and inferences are visibly separated.
  • Check which parts are concrete actions versus framing, so the user can tell what is real now.

Example Invocation

/clairvoyance get a snapshot of this [system/service/environment]. What can we observe right now through legitimate interfaces?