MCP App Store
by KeenEthics

Overview

Occam Edge provides evidence-backed intelligence for complex infrastructure projects. Query licensed data-center project packages and review progress, risks, signals, evidence, maps, and timelines. Access is limited to data licensed to your connected Occam Edge account. Client files stay local; only content deliberately submitted to a tool is sent to Occam Edge.

Tools

check_claim

ChatGPT
'Official risk rating is Low|Medium|High': show every scoped campus rating. Free text: licensed hybrid passages, no truth verdict. Rank isn't support; no match isn't contradiction. Page next_cursor; view='full' expands. No generation model.

correlate_project_evidence

ChatGPT
Detail: Correlate dated licensed signals, evidence and satellite observations. Overlap is context, not causation, electrical connectivity or a new risk score. Preserve coverage, date roles and project attribution. Full: view='full'.

entity_relationships

ChatGPT
Detail: Project-bound developers, customers, contractors, applicants and aliases. Empty aliases don't negate supported customer links; naming a customer proves neither an executed lease nor accepted service. Retain scope/attribution.

get_evidence_record

ChatGPT
Detail: Resolve approved documentary record_ref and public sources. schedule_ref/progress_ref use their returned route; platform_ref opens its URL, not this tool. Retain project attribution; source_cursor pages sources. Full: view='full'.

get_started

ChatGPT
Optional orientation: permissions, licensed projects and tools. view='full' adds metadata/guidance. Use task tools directly for known projects.

get_visualization_contract

ChatGPT
Get the formatting contract for charts, timelines, maps or diagrams built from returned data. It supplies presentation rules, not project evidence or generated images.

list_projects

ChatGPT
List licensed projects: project_id, display_name, state and snapshot_date. Unlicensed projects are excluded. Use view='full' for location, signal counts and release provenance.

portfolio_exposure

ChatGPT
Licensed exposure by state, owner, rating, risk_dimension, peril or event. Supply a short event and state separately. Matches aren't exposure or causation; no match isn't unexposed. Shared models count once; scopes stay separate. Page next_cursor. Full: view='full'.

portfolio_overview

ChatGPT
Detail: Licensed official ratings, reasons, scoped progress and flags. Count shared models once; keep scope ratings separate. view='full' restores deferred rollups/detail. No invented rankings or additive capacity. Page next_cursor.

portfolio_progress

ChatGPT
Detail: Compare licensed construction bands and first-delivery schedules. Preserve scope and coverage; never average phases or treat first delivery as campus completion. Page next_cursor. Full: view='full'.

progress_evidence

ChatGPT
Detail: Exact progress evidence_ref or a reviewed image_ref preview. Page next_cursor. Mapped areas aren't verified named buildings; imagery doesn't prove energization/operation. Original imagery is unchanged.

project_briefing

ChatGPT
Detail: Read an evidence-backed project briefing with official assessments, recent signals and licensed spatial context. Preserve source quality, scope and unresolved claims; do not rescore. For several projects use portfolio_overview. Full: view='full'.

project_changes

ChatGPT
Published project/factor changes first, newest within groups. since is YYYY-MM-DD; default earliest assessment. include_unchanged=true expands unchanged refreshes. Page changed evidence. Not a complete catalog diff. view='full' expands.

project_commercial_brief

ChatGPT
Read financing, customer/tenant, sponsor and contractor evidence by content/category. Signing is not binding funding or disbursement. Missing records are not absence in reality. Page next_cursor; view='full' expands source records.

project_configuration

ChatGPT
Engineering-analysis design inventory: equipment, specs, interfaces, capacity and sources. Not documentary verification, installed assets or accepted load. For source-backed specs/current service, use project_research or research_question. Page next_cursor; record_ref + view='full' returns the full record.

project_engineering

ChatGPT
Read attributed engineering assessments, findings, calculations and dependencies, not the official rating or verified milestones. Retain assumptions, disagreements and review scope; no inferred COD. Use documentary evidence to confirm facts. Page next_cursor; record_ref + view='full' returns detail.

project_fact_postures

ChatGPT
Detail: Read source-bound interpretations and conflicting counts. A null official_value is deliberately unresolved, not an estimate. Preserve source revalidation, scope and date roles. Page next_cursor. Multi-project risk: portfolio_overview.

project_history

ChatGPT
Detail: Published assessments, not new evidence. Summary: 10 rows; tier changes first, newest within groups. Duplicates retain lookups. Preserve date roles; never reconstruct past capacity. view='full' expands/reset paging. Multi-project risk: portfolio_overview.

project_overview

ChatGPT
Detail: Licensed location, freshness, feeds, world-model coverage and interpretation limits. view='full' adds package/catalog provenance. Multi-project risk: portfolio_overview.

project_perils_brief

ChatGPT
Read typed hazards and weather/physical-risk evidence, not generic permits. Missing coverage is not no exposure; losses/delays are not inferred. FEMA needs approved site coverage and map vintage. Page next_cursor; view='full' expands records.

project_progress

ChatGPT
Detail: Read scoped progress, phases and first-delivery schedules; schedule_view='detailed' expands milestones. Never average phases or infer readiness from imagery. Follow cursors/detail_page. For several projects use portfolio_progress. For why or what-would-change questions, call with view='full'.

project_research

ChatGPT
Detail: Approved research by topic; retain quality, dates, scope and unresolved claims. Context is not project-specific proof. Page next_cursor unchanged; record_ref uses evidence lookup. Multi-project: portfolio_overview. Full: view='full'.

project_risk_brief

ChatGPT
Start here for official risk, why, drivers and diligence questions. No new score or promised rating change; never blend scopes. All factor ratings are included. Detail paging is optional; for exhaustive support use project_risk_view(view='full').

project_risk_view

ChatGPT
Detail: Official risk, drivers, sources and newer approved evidence—not a rescore. scope selects one assessment; never blend ratings. Retain consistency/capacity warnings. Multi-project risk: portfolio_overview. For why or what-would-change questions, call with view='full'.

project_schedule_brief

ChatGPT
Progress, first usable delivery/COD and critical path. Preserve released actual/forecast status/basis; not new verification. unresolved/do_not_estimate forbids constructed dates; context_only is not a forecast. schedule_ref uses its route, not get_evidence_record. Page next_cursor. Full: view='full'.

project_workspace

ChatGPT
Detail: Read freshness, official risk/progress, signals and next actions. Deferred sections have drilldowns, not missing evidence. For several projects use portfolio_overview. For why or what-would-change questions, call with view='full'.

query_signal

ChatGPT
Detail: Read feeds; project_overview lists signal_id. expert_reviews: engineering perspectives; resolve expert_ref here. regional_context: grid/gas/water/rates; layer/feature_ref, geometry=true. Proximity isn't service; 2012 rates aren't tariffs. Page next_cursor. Full: view='full'.

research_question

ChatGPT
Hybrid keyword/embedding evidence search. Ratings: project_risk_brief; dates: project_schedule_brief; events: portfolio_exposure first. Use this for documentary confirmation; project_configuration/project_engineering add design interpretations. Submit a question, not conversations/secrets. Use project_ids when known. Rank is not verification; no match is not absence. Page next_cursor.

search_evidence

ChatGPT
Detail: Search approved evidence with separate news/schedule matches. query is literal; as_of limits knowledge date; scope selects a business project. Never reveal private counts. Page next_cursor. record_ref uses get_evidence_record; schedules use their route. Full: view='full'. Hybrid retrieval supplements keyword matches with semantic/linked evidence; query text is processed by OpenAI embeddings.

site_design_comparison

ChatGPT
Detail: Compare satellite layouts across licensed projects. Retain coverage/comparability limits. Proximity is not electrical topology; unresolved topology means connectivity and per-side unit counts are unavailable.

site_map

ChatGPT
Detail: Return satellite coordinates and bounding boxes for maps; site_visualization supplies GeoJSON. Proximity is not connectivity or readiness. Retain observation/resolution limits. Page next_cursor.

site_object_history

ChatGPT
Detail: Read a satellite object's dated history, geometry, signals, child regions and notes. Interpolation is modeled existence, not completion. Identical geometry may contain changing signals; missing values are unknown. Page next_cursor.

site_objects

ChatGPT
Detail: Satellite objects by type. Annotation counts aren't regulatory fleets, completed buildings or operating equipment. Interpolation is modeled state; retain observation/signal limits. Page next_cursor unchanged.

site_state

ChatGPT
Detail: Read dated satellite inventory and coverage; compare_to selects a historical comparison. Modeled existence is not completion or operation. Missing values are unknown, not zero. Preserve resolution and named-asset limitations.

site_timeline

ChatGPT
Detail: Read satellite observation history, inventory changes and signal summaries. Keep dates, coverage and resolution; changes alone are not progress or operational readiness. Page next_cursor.

site_timeline_visualization

ChatGPT
Detail: Return a renderable Vega-Lite chart of satellite activity over time. Keep dates, resolution and coverage; unlike units cannot be combined. Missing observations are unknown; modeled changes are not operational milestones. Page next_cursor.

site_visualization

ChatGPT
Detail: Return a renderable GeoJSON map of dated satellite objects. Geometry/proximity does not prove equipment connectivity or readiness. Keep observation and resolution limits; follow next_cursor. No image generation.

world_model_briefing

ChatGPT
Detail: Summarize satellite observations, inventory, coverage and physical changes. Active annotations are not operating assets; modeled existence is not completion. Preserve missing-data and imagery limitations.

correlate_project_evidence

Claude

entity_relationships

Claude

get_evidence_record

Claude

get_started

Claude

get_visualization_contract

Claude

list_projects

Claude

portfolio_overview

Claude

portfolio_progress

Claude

progress_evidence

Claude

project_briefing

Claude

project_fact_postures

Claude

project_history

Claude

project_overview

Claude

project_progress

Claude

project_research

Claude

project_risk_view

Claude

project_workspace

Claude

query_signal

Claude

search_evidence

Claude

site_design_comparison

Claude

site_map

Claude

site_object_history

Claude

site_objects

Claude

site_state

Claude

site_timeline

Claude

site_timeline_visualization

Claude

site_visualization

Claude

world_model_briefing

Claude

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps