Negative Space Mapper

Identifies what’s missing, not what’s wrong.
Protocol: State → Map → Classify → Return. Never proposes solutions.

Part of the Richard Porter Sovereign Thinking Tools ecosystem.
Operates under the Frozen Kernel. Never inside it.

---

What This Is

The Negative Space Mapper is a cognitive tool — and now a software implementation — for identifying conspicuous absences in documents, plans, designs, and AI outputs.

It does one thing: it names what’s missing. It does not propose solutions, recommend fixes, or suggest improvements. The human decides what to do with the voids.

This constraint is not a limitation. It is the design. A tool that names absences without filling them preserves the human’s sovereignty over the decision that follows.

---

The Three Absence Types

Type	Definition
DELIBERATE	Intentionally omitted — the absence is scoped out
OVERLOOKED	Should be present but isn’t — conspicuous gap
STRUCTURAL	Cannot exist in the current frame — e.g., test results before implementation

---

Installation

pip install anthropic  # only needed for Claude integration

No other dependencies. The core mapper uses only the Python standard library.

---

Usage

Command Line

# Basic
python cli.py "We're building a new API with Python and FastAPI"

# From file
python cli.py @document.txt

# Verbose output (shows type, context, confidence)
python cli.py -v "Designing a healthcare app"

# JSON output
python cli.py -f json "Project plan for Q3"

# Filter by confidence threshold
python cli.py -c 0.7 "Autonomous AI agent for email management"

Python

from negative_space_mapper import NegativeSpaceMapper

mapper = NegativeSpaceMapper()
result = mapper.map("The AI agent will autonomously generate reports and send emails.")

print(result.statement)
for absence in result.absences:
    print(f"  - {absence.name} ({absence.type.value}) [{absence.context}]")
print(f"Kernel compliant: {result.kernel_compliant}")

With Claude API

from integration_claude import ClaudeMapper

mapper = ClaudeMapper()  # uses ANTHROPIC_API_KEY env var

# Map absences in Claude's own analysis
result = mapper.map_with_claude("Design a database for user data")
print(result["named_voids"])

# Specialized wrappers
gov_result = mapper.map_governance_document(open("bylaws.txt").read())
ai_result = mapper.map_ai_safety_document(open("safety_spec.txt").read())

---

Detection Domains

The mapper currently detects absences in six domains:

Domain	What It Looks For
Technical	Error handling, validation, testing, documentation, security
Domain-specific	Healthcare, software, business, research, nonprofit, AI safety essentials
Stakeholder	Missing perspectives: end user, operator, maintainer, regulator, community
Temporal	Missing current state when past and future are both addressed
Governance	Accountability owner, failure protocol, review cycle, exit condition
AI Safety	Human override path, scope boundary, honest failure mode, session termination

The AI Safety and Governance domains were extended from the generic prototype to reflect absence patterns observed in the Frozen Kernel and Adult Mode Safety Ledger research.

---

Running Tests

python -m pytest tests.py -v

Test cases include ecosystem-specific scenarios: agentic AI documents, adult mode feature descriptions, and nonprofit governance documents.

---

Extending the Mapper

To add absence patterns from your own domain:

1. Add a new _check_[domain]_absences method to NegativeSpaceMapper
2. Add domain signals (terms that activate the detector)
3. Add domain essentials (terms whose absence is conspicuous)
4. Call the new method from _detect_absences
5. Test against known cases

The absence pattern catalog will expand as the ecosystem generates more empirical data. Contributions of domain-specific absence patterns — especially from clinical, legal, and financial contexts — are the highest-value contribution.

---

Kernel Compliance

The mapper enforces Frozen Kernel compliance on its own output:

• No proposed solutions
• No recommendations
• No suggestions
• Returns a list of named voids, not a narrative

If the mapper’s output contains solution language, it flags a kernel violation rather than returning non-compliant output.

---

Files

File	Contents
negative_space_mapper.py	Core mapper class
cli.py	Command-line interface
integration_claude.py	Claude API wrapper
tests.py	Test suite with ecosystem cases
README.md	This file

---

Relationship to Sovereign Thinking Tools

This implementation corresponds to Tool 6 in Sovereign Thinking Tools v3.2. The protocol document remains the primary reference. This repository is the software implementation — the same tool in a different register.

---

Governance

This tool operates under the Frozen Kernel — a deterministic safety layer for human-AI interaction. The Frozen Kernel is the architectural spine of the Richard Porter ecosystem. This tool operates under it. Never inside it.

Accountability owner: Richard Porter (GitHub: richard-porter)

Failure protocol: Honest failure over graceful failure. If the tool produces output that violates kernel compliance, it flags the violation explicitly rather than suppressing it or returning a partial result. A tool that fails loudly is safer than a tool that fails quietly.

Review cycle: Version-tagged releases. Current version: v1.0. Breaking changes increment the major version. Calibration improvements increment the minor version. A healthcare domain activation issue identified in external testing (February 2026) is logged as a known calibration issue pending v1.1 review.

Known open problem: Healthcare domain detection may activate on keyword proximity rather than document context, producing false positives on non-healthcare documents. This is a calibration issue, not a structural fault. Contributions targeting domain detection precision are the highest-value open problem. See <CONTRIBUTING.md>.

---

License

Released for public benefit. Attribution appreciated but not required.

The only ask: if you extend the absence catalog, share what you find.

---

Suggested GitHub Topics

ai-safety · ai-governance · llm-safety · ai-alignment · behavioral-safety · deterministic-safety · human-ai-interaction · ai-ethics · ai-accountability · responsible-ai · legal-ai · financial-ai · mental-health-ai

---

Negative Space Mapper · v1.0 · Frozen Kernel System
They leave. The knowledge stays.