System map
Major components, information/data flow, agent/tool boundaries, and the public/protected boundary.
Φ RESEARCH SYSTEMS · EXTERNAL RECON
One bounded, read-only recon of an AI or research system — three highest-impact findings, evidence, and an exact repair blueprint.
SYSTEM → EVIDENCE → REPAIR
01 · OBJECT
One bounded system. One read-only investigation.
The scope is fixed before work begins. If the system cannot fit one bounded recon object, BHRIGU narrows or declines it before agreement.
02 · DELIVERABLE
Major components, information/data flow, agent/tool boundaries, and the public/protected boundary.
Ranked by systemic impact, not by the number of minor defects.
URLs, repositories/files, observable behavior, contracts/documents, and reproducible public evidence where available.
What must change, where, the required invariant, and the acceptance condition.
03 · FIT
04 · BOUNDARY
05 · PROOF
Inspect existing systems and artifacts. These are proof surfaces, not testimonials or award claims.
Public identity, agent/MCP/context-memory/evaluation work, and protected-IP boundaries.
Inspect public proof ↗Bounded read-only agent capability with MCP, deterministic verification, explicit failure behavior, and no trading authority.
Inspect public proof ↗Live evidence-linked Bitcoin research corridor with explicit source and causality boundaries.
Inspect public proof →Bounded read-only research-agent evidence. No award or win is implied.
Inspect public proof ↗06 · AGENT ENTRY
External agents can send a minimal qualification object directly. This is an agent-compatible transport contract, not a claim of unattended automatic processing.
TO
bhrigu-revenue@agentmail.toSUBJECT
[Φ RECON] <system_name>Send only what is needed to qualify the object
{
"request_type": "phi_external_systems_recon",
"version": "1",
"system_name": "",
"system_url": "",
"system_type": "agent|mcp|rag|knowledge|research|conversational_ai|bitcoin_crypto|other",
"problem": "",
"primary_decision": "",
"evidence_urls": [],
"constraints": [
"read_only",
"no_credentials",
"no_production_mutation"
],
"preferred_language": "en|ru",
"reply_contact": ""
}Do not send
Expected qualification response
STATUS=QUALIFIED | NEEDS_SCOPE | DECLINED OBJECT=one bounded system PRICE=USD 300 DELIVERABLE=system map + 3 findings + evidence + repair blueprint IMPLEMENTATION=NO NEXT=scope agreement
07 · HUMAN ENTRY
No registration. No account. No questionnaire. No public intake database.
Start the reconΦ External Systems Recon System: URL: What is not working / what should be checked: Primary decision this recon should support: Relevant public evidence: Language: EN
08 · PROCESS
QUALIFY → BOUND SCOPE → AGREEMENT → PAYMENT → WORK → REPORT
09 · PRIVACY / PROTECTED IP
First contact may include public URLs, public repositories, public documentation, a problem statement, and authorized non-secret context. If private evidence becomes essential, the process stops until a separate authorized scope exists.
The report may expose observations, evidence, findings, repair blueprint, and acceptance criteria. ORION internals, private prompts/planners/evaluators/corpora, unpublished research, credentials, and patent-sensitive mechanisms remain protected.
10 · SUPPORT SEPARATION
The existing /support corridor remains voluntary research support and does not buy access, priority, ownership, or this paid recon.
/support →11 · START
Start with public evidence and the decision this recon must support.