Discoverability & DX

CLI/docs release closing the "Roadmap: Discoverability & DX" milestone (#254-258). Zero new dependencies.

dd-agents diff <run_a> <run_b> (#257)

New standalone command to diff findings between any two run directories — not just incremental-mode runs. Reuses ReportDiffBuilder.build_diff (same engine that powers the Run_Diff Excel sheet), so there's no second diff implementation to drift. Supports --json for scripting.

--json on health, memo, search (#256)

Extends the existing version/doctor/assess/validate/query --json convention to the remaining commands, for CI/scripting parity across the CLI surface.

Vector-search advisory hint (#255)

dd-agents run --verbose now logs an advisory warning once a data room crosses ~500 documents without dd-agents[vector] installed — reusing the file count file_discovery already computed, so it costs no extra filesystem walk. Advisory only; never blocks the pipeline. Docs now cross-reference the threshold from Getting Started and Running the Pipeline.

Documentation gaps closed (#254, #258)

• Documented two already-shipped features that had zero user-facing mentions: audio/video transcription (extraction/transcribe.py) and the clause library (reporting/clause_library.py).
• ContractKnowledgeGraph (reasoning/contract_graph.py) is now explicitly documented as an experimental, non-wired reasoning primitive (decision recorded in docs/knowledge-architecture.md) rather than silently unused.

Quality

• New permanent regression gate: scripts/discoverability-dx-harness.sh (lint, SAST, multi-instance isolation, dead-code diff, unit/integration + mypy --strict, E2E) — 6/6 passing.
• Live-tested against a real AWS Bedrock endpoint and real (Project Atlas) pipeline output, beyond the fixture-based E2E suite.
• 4444+ unit/integration tests pass; mypy --strict and ruff clean.

Install

pip install dd-agents==1.18.0

Docker

docker pull zoharbabin/due-diligence-agents:1.18.0

Full Changelog: v1.17.0...v1.18.0

Changes since v1.16.0

• v1.17.0 — doctor --config + cost reader + --json parity + provenance/KB (#239, #246, #241, #248) (#250)
• Update Homebrew formula to 1.16.0

Install

pip install dd-agents==1.17.0

Docker

docker pull zoharbabin/due-diligence-agents:1.17.0

What's Changed

• v1.17.0 — doctor --config + cost reader + --json parity + provenance/KB (#239, #246, #241, #248) by @zoharbabin in #250

Full Changelog: v1.16.0...v1.17.0

Changes since v1.15.0

• v1.16.0 — formula→findings + HTML/PDF parity + vision-cost accounting (#245, #244, #247) (#249)
• Update Homebrew formula to 1.15.0

Install

pip install dd-agents==1.16.0

Docker

docker pull zoharbabin/due-diligence-agents:1.16.0

What's Changed

• v1.16.0 — formula→findings + HTML/PDF parity + vision-cost accounting (#245, #244, #247) by @zoharbabin in #249

Full Changelog: v1.15.0...v1.16.0

Changes since v1.14.0

• v1.15.0 — completeness/integrity in reports + per-provider cost audit (#238, #240, #242) (#243)
• Update Homebrew formula to 1.14.0

Install

pip install dd-agents==1.15.0

Docker

docker pull zoharbabin/due-diligence-agents:1.15.0

What's Changed

• v1.15.0 — completeness/integrity in reports + per-provider cost audit (#238, #240, #242) by @zoharbabin in #243

Full Changelog: v1.14.0...v1.15.0

Changes since v1.13.0

• v1.14.0 — per-agent provider routing (#233) + VDR overrides & config docs (#236) (#237)
• Update Homebrew formula to 1.13.0

Install

pip install dd-agents==1.14.0

Docker

docker pull zoharbabin/due-diligence-agents:1.14.0

What's Changed

• v1.14.0 — per-agent provider routing (#233) + VDR overrides & config docs (#236) by @zoharbabin in #237

Full Changelog: v1.13.0...v1.14.0

Changes since v1.12.0

• v1.13.0 — Excel formula audit, VDR + request-list, IC memo, cost-by-model, gateway CI (#235)
• Update Homebrew formula to 1.12.0

Install

pip install dd-agents==1.13.0

Docker

docker pull zoharbabin/due-diligence-agents:1.13.0

What's Changed

• v1.13.0 — Excel formula audit, VDR + request-list, IC memo, cost-by-model, gateway CI by @zoharbabin in #235

Full Changelog: v1.12.0...v1.13.0

Changes since v1.11.0

• chore(release): bump version to 1.12.0 (#230)
• docs: full audit — delete legacy orphan, refresh marketing, fix broken links (#229)
• fix+verify: 100% cross-provider flow coverage (search robustness, model attribution, coverage matrix) (#228)
• fix(search): schema-safe column keys so structured search works on Bedrock/gateways (#227)
• feat(llm): make model/provider-agnosticism work end-to-end + auditable (#226)
• feat(llm): one provider/model-agnostic seam across all LLM call sites (#225)
• fix: honest model/provider-agnosticism — example default + docs (#224)
• fix(examples/agno-bindu): make the LLM provider configurable (provider-agnostic) (#223)
• Add an example: query a dd-agents report as a Bindu A2A agent (agno) (#220)
• Update Homebrew formula to 1.11.0

Install

pip install dd-agents==1.12.0

Docker

docker pull zoharbabin/due-diligence-agents:1.12.0

What's Changed

• Add an example: query a dd-agents report as a Bindu A2A agent (agno) by @raahulrahl in #220
• fix(examples/agno-bindu): make the LLM provider configurable (provider-agnostic) by @zoharbabin in #223
• fix: honest model/provider-agnosticism — example default + docs by @zoharbabin in #224
• feat(llm): one provider/model-agnostic seam across all LLM call sites by @zoharbabin in #225
• feat(llm): whole-system model/provider-agnosticism — fix silent CLI no-op, add audit receipt + doctor by @zoharbabin in #226
• fix(search): schema-safe column keys so structured search works on Bedrock/gateways by @zoharbabin in #227
• fix+verify: 100% cross-provider flow coverage (search robustness, model attribution, coverage matrix) by @zoharbabin in #228
• docs: full audit — delete legacy orphan, refresh marketing with agnosticism, fix broken links by @zoharbabin in #229
• chore(release): bump version to 1.12.0 by @zoharbabin in #230

New Contributors

• @raahulrahl made their first contribution in #220

Full Changelog: v1.11.0...v1.12.0

Changes since v1.10.1

• release: v1.11.0 — report-polish epic #201 + Bedrock e2e fixes
• Fix 3 live-Bedrock e2e issues: search on Bedrock, knowledge/chronicle path, quick-scan docs (#216, #217, #218) (#219)
• Report polish: finish HTML Report Quality Overhaul (epic #201) (#215)
• docs: add 'What dd-agents does NOT do' section (#213)
• docs(marketing): add iris logo mark to OG social-preview
• docs(marketing): refresh OG social-preview to new positioning + add distribution channel map
• ci(release): fix Docker Hub description 403 — bump action to v5 + document PAT scope
• Update Homebrew formula to 1.10.1

Install

pip install dd-agents==1.11.0

Docker

docker pull zoharbabin/due-diligence-agents:1.11.0

What's Changed

• docs: add 'What dd-agents does NOT do' section by @teja007-cyber in #213
• Report polish: finish HTML Report Quality Overhaul (epic #201) by @zoharbabin in #215
• Fix 3 live-Bedrock e2e issues: search on Bedrock, knowledge/chronicle path, quick-scan docs (#216, #217, #218) by @zoharbabin in #219

New Contributors

• @teja007-cyber made their first contribution in #213

Full Changelog: v1.10.1...v1.11.0

Changes since v1.10.0

• v1.10.1: golden sample consolidation — Project Atlas everywhere
• chore: purge all NovaBridge-era names — zero legacy repo-wide
• test+marketing: run e2e against Atlas; refresh screenshots; purge NovaBridge
• refactor: make Project Atlas the single golden sample everywhere
• marketing: Project Atlas demo deal + launch assets (synthetic, verified)
• docs: roll out cross-domain positioning + tagline repo-wide; add OG/JSON-LD meta
• docs(readme): add a named roster of all 13 agents + deep-dive lenses
• ci(release): fix Docker Hub short-description (13 AI agents, under 100 bytes)
• Update Homebrew formula to 1.10.0

Install

pip install dd-agents==1.10.1

Docker

docker pull zoharbabin/due-diligence-agents:1.10.1

Full Changelog: v1.10.0...v1.10.1

Changes since v1.9.0

• v1.10.0: trust-layer hardening + eval-robustness overhaul
• test(evals): size F1 no-regression tolerance to real variance (0.05 -> 0.15)
• test(evals): authoritative median-of-3 baseline — all 9 specialists pass live
• test(evals): treat agent timeout/non-success as INCONCLUSIVE, not 0.0 recall
• feat(agents): HR specialist surfaces each workforce risk as a distinct finding
• test(evals): capacity-K (K=2) recall matching — credit consolidation, cap stuffing
• test(evals): close 3 masking holes found by competitive audit of fa66531
• test(evals): make agent evals reliable and cover all 9 specialists (no masking)
• test(trust): cover engine wiring for tamper injection + numerical-audit text_dir
• docs(devto): update trust-layer article for shipped deterministic gates
• feat(trust): wire deterministic tamper detection + add quote-fidelity gate
• content: add dev.to article on the trust layer (audited + verified)
• brand: add iris logo — README header + MkDocs theme logo/favicon
• docs: add "How the Agents Work" — an audit-first agent anatomy tour
• Update Homebrew formula to 1.9.0

Install

pip install dd-agents==1.10.0

Docker

docker pull zoharbabin/due-diligence-agents:1.10.0

Full Changelog: v1.9.0...v1.10.0