Every .org file carries a #+filetags: line. This document is the canonical inventory of every tag in use, organised into named groups. It is the source of truth for compass lint tag validation. See Document Types for the #+filetags: format rules and the ancestor-slug invariant.

Tags added to the codebase without a corresponding entry here will be flagged by compass lint. To introduce a new tag: add an entry in the appropriate group, then use it. Story-slug tags (see §Story and ancestry slugs below) are exempt — they are validated by pattern rather than by enumeration.

Document type tags

Tags that identify the kind/role of a document. Overlap with #+type: is intentional: the tag makes the type grep-visible across the full file tree.

Tag	Description
`capture`	Inbox capture: an unrefined idea or observation.
`chapter`	A chapter included into a user manual.
`component`	Component overview: describes one `ores.*` component.
`composite`	Component fixture variant: a group-level composite component (used in test scaffolding).
`decision`	Architectural decision record.
`entity`	Codegen entity model (`#+type: ores.codegen.entity`).
`facet`	Codegen facet document: prose + tangle block for one Mustache template.
`field_group`	Codegen field-group model: a named sub-struct composed into an entity.
`fixture`	Test fixture: minimal scaffold used in codegen integration tests.
`function`	Public function definition: interface contract, inputs, outputs.
`investigation`	Time-boxed research spike: question, hypotheses, findings.
`knowledge`	General-purpose knowledge article.
`lookup_entity`	Codegen lookup-entity model: a reference-data enum table.
`manual`	Full user or developer manual (multi-chapter).
`memory`	Persistent LLM session memory.
`meta`	Meta-document: describes the documentation system itself.
`model`	Codegen model input file (JSON or org input to generator).
`module`	Codegen logical module catalogue (`#+type: ores.codegen.module`).
`plan`	Design or implementation plan (archive).
`product_identity`	Product identity: vision, mission, values at S5.
`recipe`	How-to recipe: step-by-step operational procedure.
`retrospective_release_note_audit_report`	Combined retrospective/release-note/audit report per sprint.
`runbook`	Operational runbook: multi-step procedure for a specific event.
`skill`	LLM skill: a specialised capability loaded into the agent.
`sprint`	Sprint document: mission, stories, health metrics.
`story`	Agile story: goal, acceptance criteria, tasks.
`task`	Agile task: concrete unit of work within a story.
`template`	Codegen Mustache template file (literate source).
`version`	Version document: roadmap and release record at S4.

Version tags

Tags identifying which product version a document belongs to.

Tag	Description
`v0`	Version 0: the initial development version.
`v1_0`	Version 1.0: the first stable release.
`v2`	Version 2: major rearchitecture milestone.
`v2_0`	Version 2.0: formal release of the v2 rearchitecture.

Sprint tags

Tags identifying which sprint a document belongs to. One per sprint; stories carry both their sprint tag and their version tag.

Tag	Description
`sprint_01`	Sprint 1
`sprint_02`	Sprint 2
`sprint_03`	Sprint 3
`sprint_04`	Sprint 4
`sprint_05`	Sprint 5
`sprint_06`	Sprint 6
`sprint_07`	Sprint 7
`sprint_08`	Sprint 8
`sprint_09`	Sprint 9
`sprint_10`	Sprint 10
`sprint_11`	Sprint 11
`sprint_12`	Sprint 12
`sprint_13`	Sprint 13
`sprint_14`	Sprint 14
`sprint_15`	Sprint 15
`sprint_16`	Sprint 16
`sprint_17`	Sprint 17
`sprint_18`	Sprint 18
`sprint_19`	Sprint 19
`sprint_20`	Sprint 20

Memory subtype tags

Tags that identify the subtype of a memory document (#+memory_subtype:).

Tag	Description
`feedback`	User correction or confirmation: a rule the agent should follow.
`reference`	Pointer to an external system, tool, or canonical source.
`user`	Fact about the user's role, expertise, or working style.
`project`	Fact about the work in flight or project motivation.

Index and cross-cutting tags

Tags that make a document findable via a cross-cutting dimension.

Tag	Description
`index`	This document is an index or catalogue of other documents.
`llm`	Related to LLM session tooling, skills, or memory.
`meta`	Related to the documentation system itself.
`tags`	Related to the tag taxonomy.
`frontmatter`	Related to document frontmatter conventions.
`document_types`	Related to document type definitions.

Component tags

Tags that identify which ores.* component a document primarily concerns.

Tag	Description
`codegen`	`ores.codegen`: the Python/Mustache code generator.
`compass`	`ores.compass`: the developer toolkit CLI.
`trading`	`ores.trading`: trading domain types and lifecycle.
`cli`	`ores.cli`: the command-line interface.
`qt`	Qt-based GUI client.
`shell`	`ores.shell`: the scripted NATS shell.
`http`	HTTP layer or REST API.
`nats`	NATS messaging substrate.
`compute`	Compute subsystem: grid, batches, workunits.
`database`	Database layer: PostgreSQL, schemas, migrations.
`security`	Security subsystem: authentication, authorisation, RBAC.
`tenancy`	Tenancy and multi-tenant isolation.
`emacs`	Emacs tooling: dashboards, org-roam, skills deployment.

Technology tags

Tags that identify a specific technology, tool, or framework.

Tag	Description
`api`	API component or API-layer sub-component.
`cmake`	CMake build system.
`cpp`	C++ language.
`sql`	SQL (PostgreSQL dialect).
`python`	Python language.
`git`	Git version control.
`github`	GitHub hosting and GitHub Actions CI.
`boost`	Boost C++ libraries.
`vcpkg`	vcpkg C++ package manager.
`clang`	Clang compiler or Clang-based tools.
`clang-format`	Clang-format code formatter.
`clang-tidy`	Clang-tidy static analyser.
`sccache`	sccache compiler cache.
`ninja`	Ninja build tool.
`postgres`	PostgreSQL database.
`org-roam`	org-roam knowledge graph.
`org-mode`	Emacs org-mode.
`plantuml`	PlantUML diagram tool.
`mustache`	Mustache template engine.
`latex`	LaTeX typesetting (used for PDF manual generation).

Domain and content tags

Tags that describe the business or technical domain of the content.

Tag	Description
`accounts`	User accounts and authentication.
`agile`	Agile process, sprint planning, retrospectives.
`architecture`	Software architecture documentation.
`authentication`	Authentication flows and session management.
`authorization`	Authorisation, roles, permissions.
`build`	Build system, CI, packaging.
`calendars`	Business calendars and day-count conventions.
`ci`	Continuous integration.
`currencies`	Currency reference data.
`counterparty`	Counterparty entity and management.
`deployment`	Deployment and environment provisioning.
`documentation`	Documentation tooling and conventions.
`domain`	Business domain modelling.
`entity`	Domain entity (also used as type tag).
`knowledge_graph`	org-roam knowledge graph management.
`lifecycle`	Entity or component lifecycle management.
`llm_code_review`	LLM-assisted code review tooling.
`logging`	Logging and observability.
`masd`	Model Assisted Software Development methodology.
`model`	Codegen data model.
`modelling`	Domain or system modelling.
`performance`	Performance measurement and optimisation.
`provisioning`	System provisioning workflows and wizards.
`rbac`	Role-Based Access Control.
`refdata`	Reference data (currencies, countries, calendars, etc.).
`release`	Release management and release notes.
`security`	Security features and hardening (also component tag).
`sprint`	Sprint-level document (also type tag).
`testing`	Test infrastructure, test coverage.
`trading`	Trading domain (also component tag).
`ui`	User interface design and implementation.
`ux`	User experience.

Story and ancestry slug tags

Story-slug tags and ancestry tags are not individually enumerated here because they are generated automatically from directory names and story slugs. They follow a naming pattern and are validated by pattern, not by lookup:

Version slug: v<N> or v<N>_<M> (e.g. v0, v1_0).
Sprint slug: sprint_<NN> (e.g. sprint_03, sprint_20).
Story slug: any lowercase snake_case or kebab-case string not matching a structural tag above. Generated from the story's directory name.
Housekeeping suffix: <sprint_slug>_housekeeping (e.g. sprint_03_housekeeping).

The generator at projects/ores.codegen/generate_doc.sh derives these automatically from --parent-dir and injects them into #+filetags:. A tag matching ^[a-z][a-z0-9_.-]*$ that is not in the structural tables above is assumed to be a slug or component-name tag and passes lint validation. Dots are permitted specifically to accommodate ores.* component name tags (e.g. ores.lisp, ores.org-js).

Tag inventory

Table of Contents