Agentic M&A due diligence pipeline — plan, classify, review in parallel, join the risks up, question the seller, draft the red flag report.
M&A & legal tooling
Portfolio monitoring for private equity combining Anthropic's Legal and PE plugins with a fully synthetic fund plus 26 portfolio companies. Scans and extracts from a set of reporting documents to establish covenant headroom, change-of-control consents, cross-fund vendor concentration risk, compliance boards and benchmark grids (LPAs, SHAs), with the obligatory citations.
Open the demo (email-gated — access on request) →Anonymises corporate transaction documents for a precedent database — entity discovery, preservation, Word tracked changes and an independent QA sweep, with three human checkpoints.
View walkthrough (email-gated — access on request) →Invents a complete fictional M&A data room, plants a known set of findings in it, and proves each one can be found from the documents alone — six commands, nineteen QA gates, red herrings, corrupted twins and a scorecard. Distilled from four hand-built rooms and proven on a fifth: five private synthetic VDRs of up to 4,000 documents, used to benchmark third-party DD tools, stress-test the DD pipeline, run sell-side exercises and evaluate document classifiers.
View on GitHub ↗Classifies every document in a data room by type and routes each one to the right team, writing the manifest the due diligence pipeline reads. Three hand-written signals — folder, filename and first page — and how many agree is the confidence score; Claude gives a second opinion rather than filing; anything still unsettled is listed for a human rather than guessed at. Built on SALI's LMSS taxonomy.
Private repoPlugins & tools
Persistent memory across Claude Code, claude.ai and Cowork, backed by a Notion hub of projects, session logs and durable facts. Ships as a marketplace plugin from the gregnb_ai_memory_hub repo.
View on GitHub ↗Finds the least-painful meeting window across 30 business cities in nine timezones — a DST-aware timeline scored by how many people are within working hours, with the best windows ranked and a copyable invite.
Open the demo →Home & infrastructure
Six-feed open-data stack for a Surrey parish — planning applications, flood warnings, crime, power cuts, roadworks and a weekly digest — running in Docker on a Mac Mini, syncing to Notion.
Runs locallyn8n in Docker on the Mac Mini behind Tailscale, wired to Claude Code via MCP.
Runs locallyScripts to manage and maintain a Readwise Reader backlog — dedupe, metadata triage, model classification into tag buckets, then an opt-in archive sweep. Dry-run by default; nothing is ever deleted.
Runs locally