v0.1.5

• release: publish the athena image devs can actually pull (cc4589d)
• fix(worker): check forced chain formation before user-message guard (2b3498d)
• fix(milvus): load existing collections on startup, bounded and non-fatal (6cc40ff)
• chore: remove legacy sync to bitbucket workflow (8fe713f)

v0.1.4

• fix: change Docker container user to numeric ID 1000 to satisfy k8s runAsNonRoot policy (1f99f80)
• chore(deps): bump go.mongodb.org/mongo-driver from 1.13.1 to 1.17.7 (622b160)
• chore(deps): bump go.mongodb.org/mongo-driver in /industry-benchmark (c72e9ad)
• chore(deps): bump go.mongodb.org/mongo-driver in /scripts/benchmark (87a41c3)
• chore(deps): bump go.mongodb.org/mongo-driver in /scripts/tests (04954e2)
• chore(deps): bump golang.org/x/crypto in /scripts/benchmark (b0eef37)
• Remove non-existent github.com/xdg-go/bson dependency from benchmark go.mod (528c362)
• Initial plan (4e93339)
• chore(deps): bump golang.org/x/crypto in /industry-benchmark (28e682c)
• chore(deps): bump golang.org/x/crypto in /scripts/tests (92ba971)

v0.1.3

• chore: bump version to 0.1.3 (3f49133)
• fix: commit generated protobuf files for docker build (70afe4c)

v0.1.2

• chore: consolidate release and publish workflows, bump to 0.1.2 (e0e6f2e)
• chore: remove unused bitbucket-pipelines.yml from open repo (8e8a5f5)
• chore: remove deprecated mirror to GitHub step (2efbf9e)

v0.1.1

• chore: bump version to 0.1.1 (dc74b6c)
• chore: add versioning and release workflows (40e3fc5)
• ci: sync GitHub main to Bitbucket athena-memos on every push (462a3c2)
• chore: update remaining Go 1.24 references to 1.26 in docs (d6f8bef)
• chore: align CI configs and docs with Go 1.26 (1ae1e65)
• fix: update Dockerfile builder image to Go 1.26 (d73b3f3)
• fix: patch 14 security vulnerabilities (71a5d85)
• feat: implement HealthCheck RPC with Redis + MongoDB dependency checks (24e0db7)
• benchmark(h2): H2 Anti-Tunneling experiment — PASS (70f6025)
• fix: run CalculateBridgeEntities per collection to avoid dynamic collection name in UPDATE (be2e6c6)
• chore: add comment pointing to K8s job for community re-detection (5f2c1dc)
• fix: skip Pregel if unavailable and always run CalculateBridgeEntities (7a63cbb)
• feat: add paginated ArangoDB export and igraph direct build to community detection (dd3d747)
• feat:before running in prod (b01c83a)
• feat: add H2 anti-tunneling experiment script (b9075f3)
• feat: add H4 community detection K8s job, results, and gitignore (32c06da)
• feat: add H3 graph scalability experiment scripts and results (cac6e87)
• chore: add .env.example and fix gitignore memory-server path (ed277c1)
• chore: migrate to Open Core and fix tests (c1db989)
• Refactor docs to be specific to Athena MemOS (984f9f1)
• Move root documentation into athena-memos (43ad1a1)
• Update documentation (70139a1)
• ignore datasets (e8ba09a)
• fix: wire ArangoDB connection from injected env vars with no localhost fallback (18cea9a)
• feat: hypothesis 6 (84231c0)
• add industry benchmark experiments with H1 Keplerian survival results and skip pipeline on benchmark-only changes (1e57398)
• refactor: rename module from memory-os to athena-memos (4939775)
• docs: comprehensive documentation overhaul (efa0930)
• ci: add Generate Protobuf step to pipeline before lint/test (ceec947)
• chore: untrack generated protobuf files and add to .gitignore (621a9ba)
• Revert: remove EMA confidence formula and in-process Pregel scheduler from PR #7 (5f1ea83)
• fix: preserve context_history on LTM edges and use momentum-weighted confidence so evolving facts do not silently overwrite established knowledge in the graph (309642d)
• feat: schedule Pregel community detection every 6 hours so bridge entities are identified automatically, enabling FetchContext to surface the most cross-cutting LTM nodes first (889df3f)
• fix: fetch all STM events before flush so IndexCognitiveEvents receives the full event set and older interactions are not invisible to Milvus semantic search (008df02)
• docs: pivot Analyst integration to Federated Architecture (9b47b80)
• docs: add Section 10 — DocIntel agent integration with Athena Memory OS (8e590f4)
• Wire LTMReader into SearchMemory so ArangoDB graph nodes are returned alongside MTM chain results (7546d61)
• fix: skip STM coalescing for sequential workflow steps with different step_ids (bd46ad3)
• fix: strip LLM markdown fences in entity extraction and skip chain-break for same execution (ba3af45)
• ci: move GitHub mirror step from dev to main branch pipeline (4bae07c)
• hotfix: add LTM read-path metrics, structured logging, and regression tests (84450f0)
• hotfix: implement LTM retrieval pipeline replacing not_implemented stub (5785b70)
• fix: add additionalProperties:false and complete required fields for Azure OpenAI structured output schema (945d6eb)
• ci: restore GitHub mirror step to dev pipeline (lost during merge conflict resolution) (7a663cd)
• ci: upgrade Dockerfile builder to golang:1.24-alpine to match go.mod (451335f)
• ci: upgrade Go image to 1.24 to match go.mod requirement (e5ecec5)
• fix: resolve bitbucket-pipelines.yml conflict with main (f364250)
• feat: implement Azure Blob Storage and wire init-ltm for production LTM deployment (f148dbc)
• docs: expand ArangoDB infra spec with verbose k8s and tuning guidelines (65ad080)
• docs: rewrite ArangoDB setup guide to focus on ltm_setup.go capabilities (eb2f902)
• docs,feat: add ArangoDB infra spec and community node indexes (814d52d)
• docs: Add unified architecture for Athena Memory OS (88a7cd4)
• ci: Add GitHub mirroring step to Bitbucket Pipeline (c61094b)
• feat: implement pregel-based community detection and bridge entity calculation (b6cfd46)
• feat(memory): Implement LTM pipeline with ArangoDB and Prometheus metrics (085b46f)
• fix(search): populate similarityScore and timestamp in SearchMemory response (a6ccc60)
• docs(api): add host, basePath and schemes to OpenAPI spec for Postman import (ad6d6e0)
• feat(stm): Complete STM architecture implementation and cleanup (dd2a4ba)
• fix:Removed the erroneous bson field (10d4cc8)
• Fix gitignore and untrack wrongly committed generated files (128aba7)
• Add entity extraction to MTM formation and structured LTM triples (3e0b4dc)
• P6: make SearchMemory user-scoped across all agents/services (6dc066d)
• feat(memory): Implement MTM garbage collector (3be8c6f)
• P5: dual-write STM events to MongoDB for durability (4a03b63)
• feat(memory): Enhance MTM formation and STM management (2f51de5)
• P4: metadata filters on SearchMemory + MongoDB indexes on cognitive_chains (89263f4)
• P3: two-phase MTM retrieval in GetContext (81da228)
• fix: populate MTM chain topic field via LLM topic analysis (557411e)
• feat: refactor memory retention to Ebbinghaus model and add vendor-agnostic LLM support (ca68cb0)
• chore: add bitbucket.org/ to gitignore (82044fe)
• chore: untrack build artifacts and generated files caught in git cache (4972f52)
• feat:Expand API for all event types (thought/action/observation) (e6d56ad)
• fix:Ensure chains form cause everything downstream (heat scoring, promotion, search) depends on having well-formed chains. (65ccb25)
• feat: start background worker and promoter scheduler on server boot (1491e5d)
• feat: implement GetContext, CreateSession, and SearchMemory endpoints (d6ad4f8)
• fix: correct BSON field name in promoter query (2982a9e)
• fix: correct BSON field names in all MongoDB queries (d2ea166)
• fix: generate ChainID for MTM formation + fix BSON field names in search (135f820)
• fix: worker topic detection now compares user messages across turns (5f0e414)
• fix: make Milvus vector dimension configurable via MILVUS_VECTOR_DIMENSION (3c4c6d2)
• fix: replace tanh heat scoring with normalized scoring (a9a2a4a)
• gitignore: exclude compiled memory-server binary (c309e91)
• redis fix and removing mock variables (61514a5)
• refactor(memory): cleanup broken files, improve STM/MTM implementation and add tests (02681ed)
• feat: update pipeline to deploy to prod env-container apps on main branch (a7e47ef)
• updated read me file (098b113)
• feat(queue): Implement multi-tenant Index Queue architecture 2 3 Refactors the task queueing system from a single global queue to a scalable, multi-tenant "queue of queues" pattern. This change addresses the "Noisy Neighbor" problem, ensuring that work from one agent cannot block or delay work from another. 4 5 Key changes: 6 - Producer (server.go): The API server now acts as the producer. It pushes new tasks to a private, scoped queue for each agent and then adds a pointer to that scoped queue onto a global work index. 7 - Consumer (worker.go): The background worker now acts as the consumer. It performs a two-step dequeue, first popping a scoped queue name from the global index, and then popping the actual task from that scoped queue. 8 - task_queue.go: This file has been simplified to a model/helper package, defining the queue structures but abstracting away the enqueue/dequeue logic. 9 10 Additionally, all relevant unit tests for the server, worker, and task queue have been updated to reflect this new architecture. This includes fixing mock implementations and rewriting test logic to validate the new two-step enqueue/dequeue process. (a4c725a)
• code clean up (ee87e29)
• tested application of STM (5542823)
• Subject: feat(memory): Architect STM for Cognitive Chains & Hot/Cold PathWhat We ChangedThis commit introduces a fundamental, end-to-end architectural refactor of the Athena MemOs Short-Term Memory (STM) system. The system has been evolved from a simple "chat logger" into a high-performance, robust "cognitive logger."The key changes are:1. New Data Model (dialogue.go):Replaced the old $Q+R DialoguePage model with a new, granular CognitiveEvent struct. This new model includes Role ("user", "agent") and Type ("message", "thought", "action", "observation").2. Hot/Cold Path Architecture:Hot Path (stm_cache.go): Refactored to be a "fire-and-forget" service. Its only job is to LPUSH a new CognitiveEvent to a Redis list and return immediately, ensuring zero blocking on the main thread.Cold Path (worker.go): A new asynchronous worker pool now handles all processing.3. Smart Trigger (server.go / task_queue.go):Replaced the old "heavy" MemoryProcessingTask with a "lightweight" CognitiveChainCheckTask that only contains a UserID.Implemented the "Smart Trigger" optimization (from stm_review.md) in the API handler. The worker is now only called for user/message events, preventing queue flooding from high-frequency agent thoughts.4. Worker "Brain" Logic (worker.go):The worker now runs the "Cosine Gate" logic (fast-pass, fast-fail, LLM fallback) to detect chain breaks by reading directly from the STM cache.Implemented the "Save-Then-Trim" data-safety protocol (from stm_review.md). The worker now only LTrims the STM after the MTM save is successful, preventing data loss on persistence errors.5. Service Layer (stm_store.go):Refactored the stm_store into a "hands" service layer.Removed DetermineDialogueChain.Created ProcessMTMFormation to orchestrate the entire MTM save pipeline (summarize, embed, save to Mongo/Milvus).Updated CreateSegmentSummary to use the new rich prompt from stm_review.md, enabling it to understand [thought] and [action] events.6. Testing (n...