No description
The FidelityManager's internal pressure calculation uses its own tracked object tokens divided by window_size, which is always tiny compared to the real context. Temporarily scale window_size so the FM's pressure matches the actual API input_tokens/window ratio, triggering L0→L1→L2 degradations when context exceeds 50%. |
||
|---|---|---|
| deploy | ||
| opencode-plugin | ||
| sql | ||
| src/mnemosyne | ||
| tests | ||
| .gitignore | ||
| .python-version | ||
| ARCHITECTURE.md | ||
| docker-compose.yml | ||
| mnemosyne.toml | ||
| pyproject.toml | ||
| README.md | ||
| REFERENCES.md | ||
| ROADMAP.md | ||
| SCHEMA.md | ||
Mnemosyne
Object-addressed context memory for LLM agents. Built on Pichay's demand paging foundation.
Status
Work in progress. See ARCHITECTURE.md and ROADMAP.md for design details.