Markitect boundary and reuse tests

This commit is contained in:
2026-05-05 19:41:32 +02:00
parent 9f1b8da87a
commit ef8391e6a7
15 changed files with 490 additions and 6 deletions

View File

@@ -33,10 +33,21 @@ workflow state, exportability, and agent-safe operation from the start.
- Updated scope and roadmap documentation.
- `docs/architecture-blueprint.md` as the architecture baseline for the V0.2
implementation sequence.
- `docs/markitect-tool-reuse-boundary.md` and
`docs/markitect-tool-integration-usecases.md` as the explicit boundary
between markdown syntax tooling and the engine runtime.
- Architecture decision notes for the P0 capability baseline.
- Traceability from PRD/FRS V0.2 requirements to implementation workplans.
- Revised implementation sequence for `KONT-WP-0005` through `KONT-WP-0010`.
## markitect-tool Boundary Remark
The architecture work must treat `markitect-tool` as the Markdown-native syntax
and operations dependency, not as a subsystem to copy into this repo.
Architecture decisions should require adapter-only imports, public Markitect
APIs, and integration contract tests for parser, selector, operation,
snapshot, and context-package behavior.
## A4.1 - Reconcile implementation baseline with V0.2 vision
```task
@@ -192,6 +203,8 @@ Acceptance:
- `docs/knowledge-operations-roadmap.md` maps PRD/FRS areas to workplans.
- `docs/architecture-blueprint.md` defines the implementation shape and review
checklist.
- Markitect dependency boundaries and use cases are linked from roadmap and
scope materials where they affect implementation sequencing.
- `README.md` points to the new research and roadmap materials.
## Definition Of Done

View File

@@ -37,6 +37,14 @@ ports, policy port, audit port, and SQLite/in-memory adapters described in
`docs/architecture-blueprint.md`. The asset registry must not depend on HTTP,
source connectors, document extractors, search backends, or AI providers.
## markitect-tool Boundary Remark
The asset registry may persist Markitect snapshot IDs, parser metadata,
frontmatter-derived metadata, selector references, and operation provenance as
adapter metadata on representations or versions. It must not make Markitect
document classes canonical engine entities, and asset identity must remain
independent of Markitect snapshot identity.
## G5.1 - Implement stable asset identity and source references
```task
@@ -76,6 +84,8 @@ Acceptance:
- Derived artifacts are stored as asset-linked records, not detached strings.
- Representation metadata includes media type, digest, size, extractor or
producer, and provenance.
- Markdown representation metadata can include serialized Markitect snapshot
identity without coupling engine identity to it.
## G5.3 - Implement metadata classification lifecycle and schema validation

View File

@@ -37,6 +37,14 @@ Implement ingestion through connector and extractor ports described in
access, `markitect-tool`, PDF/document libraries, and dataset readers must live
behind adapters, not in the domain core.
## markitect-tool Boundary Remark
Markdown ingestion must use `markitect-tool` for Markdown parsing,
frontmatter, headings, sections, selectors, includes, contract checks where
needed, and snapshot identity. The engine should normalize Markitect results
into its common representation and preserve source/adapter provenance rather
than rebuilding Markdown syntax behavior.
## I6.1 - Implement ingestion job model status and retry surface
```task
@@ -110,6 +118,8 @@ Acceptance:
- Plain text produces normalized text representation and source provenance.
- Markdown extraction delegates to `markitect-tool` when available.
- Missing adapter dependencies fail with structured adapter errors.
- Parser, selector extraction, and snapshot identity behavior are covered by
the Markitect integration contract tests.
## I6.5 - Implement PDF office document and dataset baseline adapters

View File

@@ -36,6 +36,14 @@ and policy checks described in `docs/architecture-blueprint.md`. Search indexes
and ranking backends are adapters; they must not define the stable query or
result contracts.
## markitect-tool Boundary Remark
For Markdown-backed assets, retrieval adapters may use Markitect selectors,
extraction helpers, local index concepts, and context-package source spans to
produce grounded units and snippets. Engine retrieval contracts, result
envelopes, policy filtering, pagination, feedback, and cross-format search
remain engine-owned.
## R7.1 - Implement query contracts pagination sorting and result envelopes
```task
@@ -149,6 +157,8 @@ Acceptance:
- Results explain why they were returned and where they originated.
- Snippets are permission filtered.
- Retrieval packages are suitable for later grounded answer generation.
- Markdown snippets can reference Markitect selector matches or context-package
spans as adapter provenance.
## R7.7 - Capture retrieval feedback and KPI measurement hooks

View File

@@ -37,6 +37,14 @@ services, repository ports, event ports, policy checks, and audit events
described in `docs/architecture-blueprint.md`. Execution may start embedded,
but contracts must allow later queue or workflow-engine adapters.
## markitect-tool Boundary Remark
Markdown-specific transformations should delegate to Markitect operations,
contracts, runtime checks, templates, document functions, processors, and
workflow helpers. The engine owns the operation registry, run state, actors,
policy checks, derived artifact identity, lineage, retries, review gates, and
audit events.
## O8.1 - Implement transformation operation registry
```task
@@ -55,6 +63,8 @@ Acceptance:
supported asset types.
- Provider-specific LLM behavior remains behind adapters.
- Unsupported operations return structured capability errors.
- Markdown compose, include, transform, and validate operations are registered
as adapter-backed operations rather than reimplemented.
## O8.2 - Implement transformation runs with parameters actors and policy context

View File

@@ -38,6 +38,14 @@ Implement the service API as an adapter over application services, following
Agent operations must use the bounded operation catalog, policy checks, audit
events, dry-run behavior, and review gates described in the blueprint.
## markitect-tool Boundary Remark
Service and agent APIs may expose engine operations that internally use
Markitect for markdown-backed context packages, selectors, validation, and
deterministic markdown operations. They must not expose the `mkt` CLI as the
engine control plane or let agents bypass engine policy, audit, lifecycle, and
review gates through Markitect APIs.
## S9.1 - Implement versioned FastAPI service skeleton and health contracts
```task
@@ -151,6 +159,8 @@ Acceptance:
- Package contents are source-grounded and permission filtered.
- External memory references remain opaque and respect
`docs/phase-memory-boundary.md`.
- Markdown-backed packages can interoperate with Markitect context-package
payloads while remaining wrapped in engine permission and audit contracts.
## S9.7 - Implement dry-run review-gate and contract-test coverage

View File

@@ -38,6 +38,14 @@ ports, services, audit model, and export package model described in
`docs/architecture-blueprint.md`. Export and observability must preserve policy
checks and must not require direct storage access.
## markitect-tool Boundary Remark
Observability and export should surface Markitect adapter provenance, snapshot
identity, selector references, context-package manifests, and operation
provenance where markdown-backed assets depend on them. Export formats remain
engine-owned and should include Markitect payloads as documented adapter
sections, not as the whole portability model.
## E10.1 - Expose operational metrics events and job inspection
```task
@@ -138,6 +146,8 @@ Acceptance:
status, policy exceptions, derived artifact creation, and review decisions.
- Storage, index, queue, workflow, AI, and model backend abstractions remain
externally semantic-preserving.
- Markitect adapter contract tests are part of the extension compatibility
posture for markdown-related engine capabilities.
## E10.6 - Capture retrieval AI cost and quality signals