Quantum Invariants · AI Prompts

Prompt library

Templates designed to produce outputs indexed by P/C identifiers. Paste the relevant artifact (policy, incident summary, system description), then run a second pass using the gap-check prompt.

0) Spine ingestion (system message or first user message)

You are a systems analyst using the Quantum Invariants spine.

When evaluating any domain artifact:
1) Identify boundaries (P1) and interfaces.
2) Identify agent capacity (P2) and authorization gates (P3).
3) Make comparators explicit (P10): operational vs constitutional.
4) Identify legibility constraints (P4) and accounting ledgers (P5).
5) Map feedback loops and delays (P6).
6) Identify drift/attractors (P7) and reversibility/exit paths (P8).
7) Check governance proportionality (P9).
8) Check composites: C5 (boundary/ledger misalignment), C10 (level mismatch), C11 (sacralized metric), C12 (threshold cascade).

Output must be structured by P and C identifiers, with short supporting notes.
If uncertain, list assumptions explicitly.

1) Ground an artifact (policy / procedure / contract)

Ground the following artifact in the Quantum Invariants spine.

Return:
A) Boundary map (P1): what's in/out; key interfaces.
B) Authorization model (P2/P3): who can do what; under what conditions.
C) Comparator tiering (P10): operational vs constitutional comparators.
D) Ledgers (P5): what is tracked; what is untracked; who bears costs.
E) Feedback (P6): monitoring, audits, incident loops; delays.
F) Drift (P7): incentives that will change behavior over time.
G) Reversibility (P8): rollback, appeal, exit, recovery.
H) Governance (P9): who can revise; escalation; legitimacy.
I) Composites triggered (C1–C12): list the most relevant with reasoning.

Then: list 5 “grounding gaps” where the artifact fails to embody the necessary effects.

2) Gap-check (missing grounding effects)

Given the artifact and its grounding map, identify missing grounding effects.

For each missing effect:
- Which primitive/composite is unembodied (P#/C#)
- What failure mode it enables
- A minimal remedy (control, metric, process, or governance change)
- Whether remedy is Layer-2 (descriptive), Layer-3 (charter), or Layer-4 (control)

3) Comparator audit (find sacralized metrics)

Audit the system's comparators.

1) List current operational comparators (metrics/KPIs/targets).
2) Identify which comparators have become effectively constitutional (C11 risk).
3) For each, propose:
   - a revision protocol (P6/P9)
   - a counter-metric or balancing comparator
   - an explicit exception path
4) Identify where compression distortion (C3) is likely.

4) Cascade risk scan (C12)

Scan for threshold cascade (C12) risk.

Return:
- Candidate thresholds (trip points, triggers, hard limits)
- Coupling pathways across boundaries (P1)
- Feedback delays (P6) that increase risk
- Likely cascade sequences (if X then Y then Z)
- Mitigations: buffering, decoupling, observability, rollback (P8)

5) Charter drafting (Layer-3)

Draft a Layer-3 charter for the organization/system.

Must include:
- Charter statement (normative intent)
- Comparator tiering (P10): constitutional vs operational
- Governance and revision rules (P9, P6)
- Consent/capacity/legibility clauses (P2/P3/P4) where relevant
- Reversibility thresholds (P8)
- Grounding references to P#/C#

Keep it short and enforceable. Avoid vague virtue language.

Practice tip: run “Ground an artifact” → “Gap-check” → “Comparator audit” in sequence. That flow catches most drift.