# LAKA Volumetric Intelligence System

> This library turns public company intelligence into evidence-backed commercial insight and respectful executive outreach. It combines passive OSINT, PR and brand intelligence, marketing-demand analysis, micro- and macroeconomics, competitor benchmarking, financial scenario modeling, and CRM-ready lead records.

Version 1.0, research baseline 2026-09-01. 23 documents,
268 catalogued tools and sources, 20 financial models,
32 safe query families, 172 acceptance items.

This is an operating grammar for company intelligence, not a prompt library. Every output must
preserve one chain:

  source -> observation -> claim state -> confidence -> mechanism -> financial range
  -> service fit -> smallest useful next step

## The rules that bound every run
- Passive public-source research is the default. No login bypass, exploitation, credential collection, intrusive scanning, pretexting, or contact-data guessing.
- A source is not a finding. Preserve its URL, publication/observation time, retrieval time, excerpt or structured fact, and content hash where practical.
- Keep Observed, Calculated, Inferred, Hypothesized, Unknown, and Contradicted separate.
- Never present an inferred security weakness, executive motive, revenue figure, sentiment percentage, or causal effect as observed fact.
- Dollar values are low/base/high scenarios with visible formulas and inputs. If inputs are unavailable, say not estimable and list the missing inputs.
- Respect robots.txt, source terms, API limits, privacy law, retention limits, and opt-outs. Public visibility does not automatically authorize collection or marketing use.
- Outreach uses business-relevant findings. It excludes family, home, health, protected traits, private routines, leaked credentials, and other sensitive personal details.
- The system may suggest a likely corporate email pattern, but the CRM must distinguish published, verified, pattern-only, and unknown. Never claim a guessed address is verified.
- Human review gates all high-impact claims, security findings, named-person profiles, financial estimates, and outbound messages.

## What the system will not do
- perform penetration testing or exploit validation;
- find or trade leaked passwords;
- infer protected characteristics or sensitive personal traits;
- diagnose an executive psychologically;
- scrape prohibited sources or defeat access controls;
- impersonate a person or create deceptive outreach;
- manufacture precision from industry benchmarks;
- turn correlation, sentiment, or share of voice into causal revenue claims;
- auto-send messages without jurisdictional and human review.

## Stop conditions
- the task requires bypassing authentication, rate limits, paywalls, CAPTCHAs, or source controls;
- a claim relies on stolen or sensitive data;
- the target cannot be reliably resolved;
- a person-directed action has an ambiguous recipient;
- evidence cannot support a proposed allegation;
- a security observation could materially increase risk if exposed in outreach;
- financial inputs are too speculative to produce a useful range;
- terms, law, or contracts prohibit the planned collection or use;
- the operator cannot explain why a data field is necessary.

## API
Base URL: https://intel.bowtiekreative.com
OpenAPI: https://intel.bowtiekreative.com/v1/openapi.json
Index: https://intel.bowtiekreative.com/v1

Read the doctrine:
- https://intel.bowtiekreative.com/v1/docs - the 23 source documents; add /{slug}?format=markdown for the file verbatim
- https://intel.bowtiekreative.com/v1/laka - the four-layer grammar
- https://intel.bowtiekreative.com/v1/laka/lenses - the 216-lens expansion cube, each lens carrying its permission rule
- https://intel.bowtiekreative.com/v1/collections - 45 typed lookup tables
- https://intel.bowtiekreative.com/v1/phases - the nine audit phases
- https://intel.bowtiekreative.com/v1/workflows - W0 to W12
- https://intel.bowtiekreative.com/v1/agents - the 14 agent prompts, verbatim
- https://intel.bowtiekreative.com/v1/tools?library=05 - the tool and source registry
- https://intel.bowtiekreative.com/v1/financial-models - the 20 impact models
- https://intel.bowtiekreative.com/v1/guardrails?jurisdiction=CA - legal and ethical guardrails
- https://intel.bowtiekreative.com/v1/search?q= - search everything

Compute:
- POST https://intel.bowtiekreative.com/v1/score/finding - finding quality, 0-100
- POST https://intel.bowtiekreative.com/v1/score/opportunity - opportunity priority with hard gates
- POST https://intel.bowtiekreative.com/v1/score/authority - outreach eligibility
- POST https://intel.bowtiekreative.com/v1/score/lead - lead qualification band

## How the scorers behave
Weights and point ceilings are read from the corpus, so the API and the documentation cannot
drift apart. Two rules hold everywhere:

- a hard gate overrides the numeric score, never the other way round;
- a missing input comes back under `needs_input`, never defaulted to zero, because an absent
  judgement and a judgement of zero are different claims.

Send `GET https://intel.bowtiekreative.com/v1/scoring/{rubric}` first for the exact dimension keys and scales.

## Provenance
Every record carries `defined_in` naming the file, section anchor and source line it was read
from. https://intel.bowtiekreative.com/v1/integrity reports whether every typed collection still resolves.

## Boundaries
Passive public-source research is the default. This library does not perform penetration
testing, defeat access controls, trade leaked credentials, infer protected characteristics, or
auto-send outreach. It is strategic and technical guidance, not legal, investment or
penetration-testing advice. Tool entries were catalogued at the 2026-09-01 research
baseline and must be re-verified before deployment.