Get Diligence Done

imagenpm imagetests

Get Diligence Done (GDD) is the open-source agentic copilot for commercial due diligence: scope the deal, lock the taxonomy, run the modules in fresh-context agents, triangulate, red-team, tell the story. A deterministic engagement spine with open-ended analytical leaves, persistent state in markdown plus a machine lock, and a verification gate whose honest claim is consistency and traceability, not truth.

Claude Code and Codex have CLI adapters. Claude Cowork and ChatGPT Work are separate native acceptance targets; their plugin installation and persistence paths are not yet certified. See docs/design.md for the full design and docs/install.md for every install path.

SOW intake pilot (0.2.4)

The candidate adds /gdd:ingest-sow <file>, stable question IDs in QUESTIONS.md, /gdd:sow-status, and local Git checkpoints. The portable plugin offers the entry point as $gdd.

The candidate includes bounded intake corrections and still needs full synthetic acceptance before real-SOW testing. Bounded Claude Code checks do not certify Cowork. The pilot targets Cowork's repository marketplace: add Navis-Advisory/gdd through Customize → Plugins → Add → Add marketplace, then install GDD after the reviewed public candidate is available. See the pilot guide for exact candidate identification, account separation and lifecycle checks. No ZIP handoff or npm release is needed.

Install

Legacy CLI adapter. Requires Node.js >= 20. For this SOW pilot, use the native plugin package and instructions above.

npx get-diligence-done --claude --global

That projects GDD into ~/.claude, so /gdd: works from any deal folder — the tool's normal per-engagement usage pattern. Swap --claude for --codex to target Codex. The installer is a single file with no dependencies; there's nothing to build.

From source (to read the code first, or to hack on it):

Run the npx command from outside a GDD clone. Inside a clone, use the source command below; npx may resolve the local package before it is installed.

git clone https://github.com/Navis-Advisory/gdd.git
cd gdd
node bin/install.js --claude --global

npm test verifies the install round-trip and reference integrity. Full details in docs/install.md.

With the global Claude CLI adapter, GDD runs out of ~/.claude; engagement state lives in a .diligence/ folder inside each deal folder, never in the install. To update a native plugin, use the host's plugin manager. CLI updates are a separate route; see installation guidance.

Claude Code plugin. Add the marketplace with /plugin marketplace add Navis-Advisory/gdd, then open /plugin install gdd@gdd and choose user, project, or local scope. User scope makes it available across your projects on that machine.

Claude Cowork and ChatGPT Work. Native installation and persistence acceptance are pending. Do not assume a CLI installation makes the plugin available in either app. See the installation guide for current status.

Why we built this

We run commercial due diligence for a living. Pointing a general-purpose AI at a diligence question rarely fails by being obviously wrong — it fails by being confidently unverifiable: a clean-sounding market size with no traceable source, a thesis nothing ever tried to kill. That's the failure a good engagement manager exists to prevent. GDD encodes that discipline — every number cited, every market sized two independent ways, every thesis red-teamed — as a workflow, so the rigor is structural rather than hoped-for.

It's a scalpel, not an autopilot: trust the execution, verify like the EM.

The loop

scope-deal → hypothesis-tree → workplan
  → size-market / map-competitors / probe-customers / assess-moat / scan-risks
    (modules, fresh-context, independent)
  → triangulate (verification gate) → red-team → storyline

• scope-deal — interview the deal into an engagement brief and a locked taxonomy (the segment definitions everything downstream must obey).
• hypothesis-tree — decompose the investment thesis into a MECE tree of falsifiable hypotheses, one brief per module.
• workplan — lay the modules across the engagement timeline, dependency-ordered.
• modules — the analytical work, each in a fresh-context agent: size-market, map-competitors, probe-customers, assess-moat, scan-risks.
• triangulate — the verification gate: consistency and traceability checks across every finding. A FAIL blocks the storyline.
• red-team — the strongest case against the thesis; surviving kills go CONTESTED.
• storyline — synthesize the surviving findings into a pyramid- principle storyline.

Why it works

Naive AI diligence fails in three predictable ways. GDD is built around closing each one:

• Untraceable claims. Every finding carries a citation to a tiered source; the verifier fails the gate when a load-bearing number doesn't trace back to one.
• Single-source sizing. The market is sized top-down and bottom-up by two agents that never see each other's work, then reconciled — an unreconciled gap is reported and named, never averaged away.
• Unfalsified theses. A red-team stage argues the strongest case against the thesis before the story is told; a killed finding goes CONTESTED and can't carry a key-line claim.

The verifier's promise is consistency and traceability — it confirms the TAM math checks out, top-down and bottom-up reconcile, every number traces to a cited source, and the segment tree is MECE. It never claims the market is actually $X.

Layout

User engagement state lives in .diligence/ inside a deal folder (never in this repo — gitignored): ENGAGEMENT.md, TAXONOMY.md + state.json taxonomy lock, LEDGER.md, SOURCES.md, WORKPLAN.md, STATE.md, modules/*/, reports/.

Contributing

Issues are open and welcome — bug reports and feature ideas both. Pull requests are restricted to collaborators (this repo is a mirror). See CONTRIBUTING.md.

Inspiration

GDD takes its name and skeleton in analogy with GSD (MIT), whose command-workflow idea transfers cleanly beyond software development. Machinery here is written fresh against that pattern.

Cloud sessions

Claude Code cloud sessions self-install this checkout's GDD workflows through the tracked SessionStart hook. For Codex Cloud, set the repository Environment setup script to:

bash scripts/cloud-bootstrap.sh codex

The script installs GDD locally into the ephemeral clone, so the commands match the checked-out revision rather than an unrelated user-level install.

License

Apache-2.0. Copyright 2026 Navis Advisory.