- retrigger release with sr v7.1.2 Cargo.lock handling (20eacdf)
- fold init, add progress TUI, layer config (8bce2d3)
- ci: apply cargo fmt, stage Cargo.lock in releases (91c7fc8)
- drop nonexistent sr init --merge flag (babc735)
- move cargo publish into sr hooks.post_release (41a0510)
- showcase: re-record demo with measured wait durations (d9e2ad4)
- ci: drop mnemonist-index from publish step (crate no longer exists) (da0d1c1)
- ci: use checked_div for manual_checked_ops lint (rust 1.95) (a38ff94)
- ci: resolve clippy and fmt failures in bench experiments (7b35479)
- make each bench experiment independent and self-contained (85f8432)
- update Cargo.lock (c699e99)
- add --hidden, --no-gitignore, and --exclude flags for code indexing (fc87447)
- default install dir to ~/.local/bin and simplify install script (c183924)
- ci: migrate sr v4 to v7 for artifact and input support (#10) (74d0273)
- benchmarks: add reference to raw benchmark results (c632c0b)
- consolidate index, quant, evals into core (#9) (a012b66)
- migrate sr config and action to v4 (9a18321)
- sync embedded files [skip ci] (be7f035)
- release: migrate from embed-src to fsrc (86661ef)
- remove legacy .llmem directory from repo (5d07cc4)
- evals: add end-to-end LongMemEval QA harness and fix retrieval methodology (afa7c7c)
- evals: add benchmarking suite and LongMemEval evaluation (abc866d)
- core: implement candle-based embedder with hardware acceleration (2db6c9b)
- core: update examples to use CandleEmbedder (80b01b4)
- update for candle embedder and hardware acceleration (a56aa4e)
- core: update default embedding provider to candle (3b0ba67)
- bump workspace version to 0.5.2 (09aeca0)
- improve floating-point precision in test constants (4be7a51)
- deps: migrate from fastembed to candle embeddings (f301b4d)
- scripts: add benchmark data download infrastructure (f9b8c2f)
- ci: remove --allow-dirty from cargo publish (197879b)
- ci: use native ARM runner for aarch64-linux build (88a93bb)
- cli: add self-update, version subcommands; standardize --format flag (2aea8ba)
- add LICENSE and README to sub-crates for publishing compliance (94e0754)
- readme: update features and quick start guide (3333bbf)
- recall: remove server crate, add multi-signal reranker with auto-calibration (1bf8b6e)
- add linguist overrides to fix language stats (ecc00c0)
- spec: update llmem specification to v0.2.0 (#6) (46dcad7)
- index: replace tree-sitter with pluggable ChunkingStrategy (#7) (1a8e562)
- rename project from llmem to mnemonist (6eb9d11)
- cli: canonicalize root path + add mnemonist-evals crate (#5) (35d3d15)
- readme: separate install commands into individual code blocks (9993759)
- ci: drop musl targets from build matrix (1cc06c8)
- build: switch fastembed from native-tls to rustls-tls (5a8978a)
- fastembed migration, daemon mode, TUI output, and associative linking (#4) (3ca2c39)
- update sr action from v2 to v3 (3610251)
- lowercase justfile with expanded recipes and add crates.io badge (3295794)
- standardize CI/CD — clean sr.yaml, justfile recipes, embed-src sync, teasr demo (83dc92d)
- validate: add comprehensive E2E validation script (8f27e2c)
- test: redact embedded field in memorize snapshot (076c74e)
- cli: fallback to created_at when last_accessed is not set (98e1efc)
- add benchmark results to README via embed-src (bd9d6b1)
- add testing guide to README and CONTRIBUTING (a5127f2)
- rename mnemonist-cli package to mnemonist for simpler cargo install (f9be434)
- server: cache resolved project memory directory in app state (9783d32)
- add criterion benchmarks and insta snapshot tests (d08e712)
- cli: add comprehensive integration tests (22b1248)
- server: extract router builder and add comprehensive tests (e60cc66)
- deps: add testing dependencies to server and CLI (53bb556)
- core: add working memory inbox module (0d3c93f)
- core: add memory metadata and consolidation support (da5cf63)
- quant: add TurboQuant vector quantization library (8f434ed)
- train: add python training infrastructure (1afbfd3)
- centralized storage and config system (25728c6)
- JSON-first CLI, Ollama embedder, recall/learn/code commands (1682010)
- update architecture and api references for cognitive memory (aa09b36)
- readme: document cognitive CLI and memory features (83501d7)
- spec: specify cognitive memory with inbox and consolidation (8291e0b)
- quant: add TurboQuant vector quantization library documentation (0b9c431)
- skill: add mnemonist agent skill specification (cd5ffe5)
- data: add turboquant reference and dataset structure (779072a)
- readme: add install.sh as primary install method (bcf0fe4)
- add per-crate READMEs for crates.io (68af45e)
- readme: add config commands and configuration section (25f0aaa)
- cli: redesign commands for cognitive memory system (340f8ba)
- rename .ai-memory to .mnemonist (0dbf2d2)
- skills: rename ai-memory skill to mnemonist (897a913)
- workspace: add mnemonist-quant crate and dependencies (6730456)
- add install.sh for downloading release binaries (b84bd6d)
- add Justfile with build, install, test, and fmt commands (bc7dc07)
- release: migrate release workflow to use sr github action (7bb80dc)
- update memory files (481f52f)
- initial mnemonist implementation (3a39426)