It fills a critical gap in agent observability by providing a tamper-proof record of decisions, positioning it as foundational infrastructure for agent ecosystems.
Evidence quotesverbatim, from the product’s own materials
“public GitHub repository (README fetched)”— structural
“classify each call as *determinable × reversible*, then **decide** (grounded by an artifact), **assume** (safe default, logged for async review), or **escalate*”— readme
“runs on Claude Code, Codex, Gemini CLI, Cursor”— readme
“any skills-compatible agent”— readme
“It ships as an Agent Skill — pure markdown, no scripts — so it runs on Claude Code, Codex, Gemini CLI, Cursor”— readme
FAQ
What is Log-decisions?
A log-only journal that records every decision made by an AI agent, enabling engineers and researchers to trace reasoning and verify actions later.
What does Log-decisions do?
To document each step of an AI agent’s decision-making process for later review, debugging, or compliance checks.
Why does Log-decisions matter?
It fills a critical gap in agent observability by providing a tamper-proof record of decisions, positioning it as foundational infrastructure for agent ecosystems.
How much does Log-decisions cost?
Free
Is Log-decisions free?
Yes — Log-decisions has a free tier. Pricing as stated on its own page: Free
Is Log-decisions open source?
Yes — Log-decisions is open source.
What does Log-decisions integrate with?
claude-code,codex,gemini-cli,cursor
How popular is Log-decisions?
As tracked by KanonAgent: 2 upvotes (first indexed 2026-08-11).
What are the best Log-decisions alternatives?
Similar AI agents tracked by KanonAgent: ponytail, claude-mem, open-design, deer-flow, Agent-Reach, ruflo.