by taylorsatula
This is the public release of MIRA OS. Discrete memories decay through momentum loss, tools auto-configure when dropped into tools/ folder, and the system prompt composes from modular trinkets. I would like to think I've made an elegant brain-in-box. You load it and send cURL requests - it talks back, learns, and uses tools. Contributions welcome.
# Add to your Claude Code skills
git clone https://github.com/taylorsatula/mira-OSS10 months or so ago I had the idea to build a recipe generator that could incorporate my cuisine preferences. 10,000 scope creeps later MIRA is a comprehensive best-effort approximation of a continuous digital entity.
This is my TempleOS.
Mira accomplishes the end goal of continuity and recall through a blend of asynchronous conversation processing akin to REM sleep and active self-directed context window manipulation. There is one conversation thread forever. There is no functionality to 'start a new chat'. This constraint forces facing the hard questions of how to build believable persistence within a framework (forward pass transformers) that is inherently ephemeral.
When a conversation segment collapses, Mira generates a first-person summary ("I debugged the IndexError in process_batch.py...") rather than third-person ("The assistant discussed debugging..."). Early on I noticed third-person summaries created epistemic distance—Mira read them as logs about someone else, not memories of work It actually did. Absolute timestamps ("On Jan 8") replace relative time ("Yesterday") because relative time becomes a lie the moment the sun sets.
When generating a new segment summary, the model sees the previous five summaries as context. This allows the new summary to reference what came before with hazy continuity—"Building on Tuesday's API work..." or "Continuing the recipe experiments from last week...". The result is a continuously advancing narrative in the conversation history that gracefully tails off into the past. Each summary knows vaguely where it came from without carrying the full weight of everything before it.
I have painstakingly designed Memory in MIRA to require no human intervention or curation. You will never need to housekeep old memories (context rot). They unless they Earn Their Keep by being referenced in conversation or linked to via other memories. Memories are discrete synthesized information that is passively loaded into the context window through a combination of semantic similarity, memory traversal, and filtering. Mira does not manually search for memories. They are presented just-in-time after multiple levels of recall & filtering.
No comments yet. Be the first to share your thoughts!
However,, sometimes you just need big ole chunks of document-shaped text. Mira handles this aspect via the domaindoc_tool which allows Mira & You to collaborate on stable textfiles that do not decay. To mitigate token explosion in longform content MIRA is able to expand, collapse, and subsection these text blocks autonomously. When a domaindoc block is not in use MIRA 'closes the drawer' and the content of the block is no longer polluting the context window but MIRA sees a section title so it can reexpa...