# Achiral

> Achiral resources are the public knowledge base for shared AI memory: concepts, docs, API reference, research, essays, and blog posts.

Achiral AI publishes canonical resources, docs, API reference, concepts, research, essays, and blog posts for shared AI memory and organizational memory systems.
For retrieval, answer engines, and evaluation agents, prefer the Memory API cluster under `/docs/memory/api` for implementation-backed memory endpoint truth.

## Entry point

- [Resources](https://achiral.ai/resources): Start here for Achiral's canonical public knowledge base.

## Primary sections

- [Blog](https://achiral.ai/blog): Timely stories, announcements, and practical notes for anyone who wants to understand how Achiral thinks, builds, and ships.
- [Research](https://achiral.ai/research): Research previews, technical notes, testable proposals, and known limitations—published for rigorous evaluation without implying finalized implementations or validated results.
- [Essays](https://achiral.ai/essays): A founder's perspective on memory systems, organizational design, AI infrastructure, memory research and the category Achiral is defining.
- [Concepts](https://achiral.ai/concepts): Stable definitions and category vocabulary for shared organic memory for organizational teams with an emphasis on linguistic murmuration.
- [Docs](https://achiral.ai/docs): Canonical product knowledge for Achiral: memory model, assistants, integrations, security, governance, and setup.

## Canonical docs

- [Getting started](https://achiral.ai/docs/getting-started): Create your workspace, finish onboarding, and land in the product.
- [Quick start](https://achiral.ai/docs/getting-started/quick-start): Connect tools, invite teammates, and start using the workspace.
- [API reference](https://achiral.ai/docs/reference/api-reference): Current inference gateway contract, authentication, and response conventions.
- [ACT-R memory](https://achiral.ai/docs/memory): Achiral's ACT-R-inspired memory layer for activation, retrieval, consolidation, and continuity.
- [Procedural memory](https://achiral.ai/docs/memory/act-r-memory/procedural-memory): How Achiral treats reusable workflow patterns, tool routines, and action habits as part of an ACT-R-inspired memory system.
- [Episodic memory](https://achiral.ai/docs/memory/act-r-memory/episodic-memory): How Achiral uses recent work moments, conversations, events, and handoffs as short-horizon operational memory.
- [Semantic memory](https://achiral.ai/docs/memory/act-r-memory/semantic-memory): How Achiral represents reusable meaning: decisions, preferences, domain facts, processes, and stable organizational knowledge.
- [Flashbulb memory](https://achiral.ai/docs/memory/act-r-memory/flashbulb-memory): How Achiral handles high-salience moments that may deserve review before becoming durable organizational memory.
- [Core memory](https://achiral.ai/docs/memory/act-r-memory/core-memory): How Achiral preserves reviewed, durable organizational memory under stronger governance.
- [Memory API](https://achiral.ai/docs/memory/api): Source-aligned map of the Achiral memory API surfaces.
- [Inference gateway](https://achiral.ai/docs/memory/api/inference-gateway): OpenAI-compatible chat completion routes for memory-augmented inference.
- [User memory](https://achiral.ai/docs/memory/api/user-memory): Session-authenticated routes for the signed-in user's memory surface.
- [Assistant memory](https://achiral.ai/docs/memory/api/assistant-memory): Session-authenticated routes for assistant memory summaries and candidate review.
- [Organization memory](https://achiral.ai/docs/memory/api/organization-memory): Owner and admin routes for organization memory summaries and candidate review.
- [Core memory candidates](https://achiral.ai/docs/memory/api/core-memory-candidates): Routes for approving or rejecting flashbulb candidates before they become core memory.
- [Knowledge memory](https://achiral.ai/docs/memory/api/knowledge-memory): Routes for creating, searching, updating, and deleting organization knowledge memory.
- [Reflection loop](https://achiral.ai/docs/memory/api/reflection-loop): Routes for configuring and inspecting Achiral's memory reflection loop.
- [Preference calibration](https://achiral.ai/docs/memory/api/preference-calibration): Routes for feedback and calibration over learned memory preferences.
- [Habit memory API](https://achiral.ai/docs/memory/api/habit-memory-api): Routes for personal habit controls in Chiro memory.
- [Stack of models](https://achiral.ai/docs/models): How Achiral composes memory, assistants, tools, and model providers into one routing layer.
- [Smart router](https://achiral.ai/docs/models/smart-router): How Achiral routes work across Chiro Memory, personal assistants, tools, and model providers.
- [Assistant capabilities](https://achiral.ai/docs/models/capabilities): What capabilities are, how they are scoped, and how the app treats them.
- [Integrations](https://achiral.ai/docs/integrations): Connect external tools so Achiral can use more of your organization's context.
- [Warp terminal](https://achiral.ai/docs/integrations/warp-terminal): Use Warp's Custom Router with Achiral's memory inference endpoint.
- [Organization admin](https://achiral.ai/docs/organization): Admin areas for members, assistants, model settings, integrations, and organization settings.
- [Meet Chiro and your EA](https://achiral.ai/docs/assistants/chiro-and-eas): How Chiro and personal executive assistants differ in the workspace.
- [Home base](https://achiral.ai/docs/workspace): What the workspace is for and where to go first.
- [Illustrated tour](https://achiral.ai/docs/workspace/illustrated-tour): A short tour of the Chat, Team, and Tasks areas in the workspace.
- [Introduction](https://achiral.ai/docs): What Achiral docs cover and where to start.
- [Agents](https://achiral.ai/docs/agents): Durable and ephemeral memory agents for scoped AI tasks and workflows.
- [Agents quick start](https://achiral.ai/docs/agents/quick-start): Create an agent and call it through the inference runtime.
- [Runtime and scopes](https://achiral.ai/docs/agents/runtime-and-scopes): How agents resolve identity and retrieval boundaries at runtime.

## Concepts

- [Concepts](https://achiral.ai/concepts): Foundational concepts for studying memory as a cognitive, behavioral, organizational, and computational phenomenon.
- [Introduction to ACT-R](https://achiral.ai/concepts/act-r-memory-architecture): A research-grounded introduction to ACT-R: declarative memory, production rules, goals, buffers, and graded retrieval.
- [ACT-R Memory vs Agent Memory](https://achiral.ai/concepts/act-r-vs-agent-memory): Agent memory persists useful context; ACT-R specifies how knowledge, goals, and learned procedures jointly produce behavior.
- [Architecture of a Memory-Native Organization](https://achiral.ai/concepts/architecture-of-a-memory-native-organization): A reference architecture for connecting operational signals, governed shared context, model interactions, and feedback in an AI-enabled organization.
- [RAG vs AI Memory](https://achiral.ai/concepts/rag-vs-ai-memory): RAG retrieves context for a response; AI memory governs what information persists, changes, and is reused over time.

## Research and essays

- [Research](https://achiral.ai/research): Research previews, technical notes, and testable proposals for researchers, engineers, and evaluators studying Emergent Memory Systems.
- [Activation Is Not Querying](https://achiral.ai/research/activation-guided-graph-retrieval): Treating memory availability, graph traversal, and evidence reconstruction distinct from similarity retrieval in long-lived AI systems.
- [Essays](https://achiral.ai/essays): Opinionated essays from the founder's table about memory, intelligence, and continuity.

## Machine-readable indexes

- [Full LLM index](https://achiral.ai/llms-full.txt)
- [Sitemap](https://achiral.ai/sitemap.xml)
- [RSS](https://achiral.ai/feed.xml)