Agent-native investment research workspace for evidence-tracked, refreshable investment theses across equities, earnings, macro, and portfolio review.

Capture the why behind AI-generated code. Generate decision logs from Claude Code sessions for PR reviews.

Your code knows what. Lore knows why. — CLI tool that captures the why behind code changes.

An append-only ledger that records why decisions were made.

Reusable agent rules, templates, subagents, and skills for AI-assisted software delivery.

SpecFirst é um framework documental em Markdown, distribuído como template, para governar projetos tocados por humanos e agentes de IA.

Session Continuity Protocal (SCP): a lightweight protocol to prevent multi-session AI “amnesia” by persisting project state (PMP), decisions (ADR-lite), and no-silent-changes guardrails for consistent, token-efficient continuity.

An append-only reasoning log for LLM-assisted research. Record the why behind decisions — with falsifiability conditions, confidence tags, and failure-first philosophy.

Self-hosted database and API for AI project memory. Store decisions with receipts, query from any model

Capture decisions with full context, metadata, and confidence scores. Includes outcome tracking, notes, and CSV export.

Branch-aware AI decision logging for development projects — AGENTS.md-canonical, link-integrity CI, skills.sh integration.

AI assumption firewall for MCP agents — intercepts risky decisions, checks past decisions, detects contradictions, and gates human approval via Telegram or Slack.

Decode technical decisions into plain language. AI drafts the why — the engineer confirms in one sentence.

Decision-first experiment tracker for product teams (hypothesis → decision → learning).

Lightweight CLI for keeping a per-repo logbook— one append-only logbook.md in your repo captures the why behind each code choice

Open specification for tamper-evident AI decision logs. Satisfies EU AI Act Articles 12 and 19. Defines record fields, hash chaining, Ed25519 signatures, and audit trail export format.

Minimal, zero-dependency decision log for Claude Code. Markdown-only, git-trackable, three smart hooks.

Persistent decision memory for any project. Karpathy's LLM Wiki, productized.

A platform that makes delivery friction visible, measurable, and actionable — so teams can fix real root causes instead of guessing what went wrong.