Meetless
A governance layer that captures human-approved decisions and injects them into AI coding agent prompts, flagging conflicts before agents act on stale information.
🔗 Visit Meetless
Description
If you've ever used an AI coding agent for more than a day or two, you've probably run into this: you tell it something important — "we're not using that library anymore" or "always validate input on the server, not the client" — and a week later, in a different session, the agent has completely forgotten and does the opposite. Meetless is built to fix that. Think of it as a team's shared memory for working with AI agents: every important decision gets written down, approved by a real person, and then automatically fed back into the agent's instructions every time it starts a new task. If two decisions ever contradict each other, Meetless flags the conflict for a human to resolve instead of letting the agent silently pick one and run with it.
Under the hood, Meetless works by capturing project decisions in a structured store, then injecting the relevant, currently-valid ones directly into the agent's prompt context at session start. It performs conflict detection between decisions — for example, if a new rule contradicts one recorded three weeks earlier — and surfaces those conflicts for human review rather than resolving them automatically. Each decision carries a tracked authority (who approved it, when), so agents always work from human-approved, non-stale context instead of an outdated CLAUDE.md file or a prompt template nobody has updated in months. It's designed for tools like Claude Code and Codex, and works whether you're a solo builder juggling multiple long-running sessions or a team trying to keep everyone's agents aligned on the same set of ground rules.
💬 Our review
The short version: Meetless tackles a real and growing pain point — AI coding agents that lose track of decisions across sessions and confidently act on stale or contradictory instructions. It adds a human-approval and conflict-detection layer on top of the context you feed your agents, which static files simply can't do on their own.
Compared to the current default — a CLAUDE.md file, a Cursor rules file, or a hand-maintained prompt template — Meetless's real advantage isn't the concept (teams have been writing project rules into text files for years), it's the automation: static files are simple, free, and require zero setup, but they don't detect when a new instruction contradicts an old one, they don't track who approved what, and they don't stop stale guidance from being fed to the agent forever. Meetless trades that simplicity for structure — you get conflict surfacing and decision authority tracking, at the cost of another tool and workflow to maintain. Since it's currently free during the beta with no credit card required, the value equation is straightforward for teams already running into agent context drift: there's essentially no financial risk in trying it. The real risk is product longevity and whether the free beta converts into a paid model you're willing to pay for once it matures — not whether it's worth the money today, because today it doesn't cost any.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Gratuit pendant la période bêta, sans carte bancaire requise. Le modèle de tarification futur n'est pas encore communiqué.
Pros
Détection automatique des conflits entre décisions de projet
Décisions validées par un humain avant injection dans le prompt de l'agent
Traçabilité de l'autorité de décision (qui a approuvé, quand)
Gratuit pendant la bêta, sans carte bancaire requise
Cons
Modèle de tarification futur inconnu, incertitude sur la pérennité du produit
Ajoute un outil et un workflow supplémentaires par rapport à un simple fichier statique
Solution encore jeune (bêta), écosystème et intégrations à confirmer dans la durée