mirai-graph 1.0.0-rc.4 → 1.0.0

package/CHANGELOG.md

@@ -4,6 +4,148 @@ All notable changes to Mirai Graph will be documented in this file.
 
 ## [Unreleased]
 
+## [1.0.0] - 2026-08-17
+
+### Changed
+
+- Project Technology is now the stable public execution mechanism for ordinary
+  projects, skills, platforms and multi-repository systems.
+- The package, CLI and programmatic API leave release-candidate status without
+  changing the public `graph.json` schema or the accepted safety boundaries.
+
+### Evidence Boundary
+
+- Stable means the documented contracts and release checks are supported for
+  production adoption. It does not claim that every project outcome is correct
+  without project-specific goals, evidence and review.
+
+## [1.0.0-rc.11] - 2026-08-17
+
+### Fixed
+
+- The bounded architecture contract now preserves the exact human architecture
+  owner together with component, capability and operational ownership data.
+- Federation adapters can consume the public contract without reconstructing
+  owner identity from a package owner.
+
+## [1.0.0-rc.10] - 2026-08-17
+
+### Fixed
+
+- Project Technology now reads complete tracked-file inventories in large
+  repositories instead of accepting Node's default child-process buffer.
+- A regression proves that an accepted target remains discoverable after more
+  than one megabyte of earlier tracked paths.
+
+## [1.0.0-rc.9] - 2026-08-17
+
+### Fixed
+
+- `mirai-graph technology ...` now resolves an omitted repository or `.` from
+  the caller's working directory instead of the installed package directory.
+- CLI regressions cover omitted, relative and absolute repository selection.
+
+## [1.0.0-rc.8] - 2026-08-17
+
+### Added
+
+- Public `mirai.project_technology` extension contract for projects, skills,
+  platforms and multi-repository systems.
+- One CLI and programmatic API for safe inventory, task-scoped context, target
+  provider export/import, exact verification and idempotent refresh-binding.
+- Fail-closed fixtures for incomplete or tampered contracts, stale providers,
+  dependency cycles, disabled state and blocked significant work.
+
+### Changed
+
+- `skill_runtime` remains the vocabulary for skill-specific knowledge, while
+  Project Technology supplies the shared execution mechanism.
+- Local mutable state now belongs under `.mirai-graph/project-technology/`.
+
+### Safety Boundary
+
+- Bounded provider exports contain identifiers, constraints, approved scope
+  and digests only. Raw documents, private source and full graph content remain
+  with their owners.
+- Read-only diagnostics remain available when significant work is blocked.
+
+## [1.0.0-rc.7] - 2026-08-06
+
+### Added
+
+- Public `graph.json` schema `2.0.0` and separate immutable dependency-lock schema.
+- Read-only-by-default `mirai-graph migrate`, with `--apply`, backup, atomic activation, automatic rollback and idempotent repeat execution.
+- Cross-format readers for the former root workspace descriptor, inner graph manifest and public package manifest.
+
+### Changed
+
+- `init`, examples, pilots and starters now write one root `graph.json`; `graph/` contains graph data only.
+- Legacy identifiers become aliases, while known legacy metadata is preserved in namespaced extensions.
+- Validation, context, readiness and adopter workflows consume the same v2 manifest.
+
+### Compatibility Boundary
+
+- Legacy manifests remain readable for one public RC, but all writers emit v2 only.
+- Migration fails closed on identity conflicts, unknown legacy fields or missing package entrypoints.
+- Graph objects, relations and raw technology sources are not rewritten by manifest migration.
+
+## [1.0.0-rc.6] - 2026-06-27
+
+### Added
+
+- Release consolidation notes for the anti-drift / quality-control contour that
+  connects Semantic Intent Resolution, Dynamic Episode tracing, Goal Vector
+  Quality Control and Technology Quality Feedback.
+- Publication foundation package for Semantic Intent and Goal-Vector Control.
+- Publication foundation package for Observable Self-Correcting AI Workflows.
+
+### Changed
+
+- Working package version is now `1.0.0-rc.6`.
+- README, release index and evidence package now describe the consolidated
+  anti-drift contour as a review surface instead of separate isolated features.
+- Publications index now separates the two article candidates and their evidence
+  requirements.
+
+### Claim Boundary
+
+- This release candidate consolidates already validator-backed public shapes and
+  publication scaffolds.
+- It does not claim peer-reviewed effectiveness, real-world drift reduction,
+  production autonomous execution safety or authorization from traces, evidence,
+  feedback, goal vectors, semantic confidence or proposals.
+
+## [1.0.0-rc.5] - 2026-06-27
+
+### Added
+
+- Semantic Intent Resolution standard as an input-control extension contract.
+- Semantic intent schema, synthetic positive example and negative fixtures for
+  low-confidence execution and proposal-as-canonical-update violations.
+- CLI validation support for `semantic-intent-resolution`.
+- Goal Vector Quality Control standard as a quality-control extension contract.
+- Goal-vector schema, synthetic positive example and negative fixtures for
+  missing vector links, false completion, tests-as-acceptance, missing reverse
+  audit and correction-without-route violations.
+- CLI validation support for `goal-vector-quality-control`.
+
+### Changed
+
+- Working package version is now `1.0.0-rc.5`.
+- README, documentation map and release index now list Semantic Intent
+  Resolution and Goal Vector Quality Control as part of the release-candidate
+  standard surface.
+- `npm test` includes Semantic Intent Resolution and Goal Vector Quality
+  Control positive and negative checks.
+
+### Claim Boundary
+
+- This release validates the public shape and safety boundaries of Semantic
+  Intent Resolution.
+- It does not claim production readiness, automatic correct process selection,
+  live-action authorization or canonical updates from generated briefs,
+  proposals, evidence, reverse audits, goal vectors or semantic confidence.
+
 ## [1.0.0-rc.4] - 2026-06-16
 
 ### Added

package/README.md

@@ -10,7 +10,7 @@ Use it when a software project has useful knowledge spread across README files,
 docs, issues, chats, code comments and AI prompts, and you need a small
 machine-readable map that developers and AI assistants can validate and reuse.
 
-Status: public `1.0.0-rc.4` release candidate.
+Status: stable public `1.0.0` release.
 
 ## What Problem Does It Solve?
 
@@ -42,9 +42,8 @@ Start with a normal repository and use Mirai Graph to:
 
 ## Current Installation Status
 
-The repository is ready as a release candidate, but `mirai-graph@1.0.0-rc.4` is
-not published on npm yet. Until npm publication is complete, use the repository
-checkout path:
+The repository is ready as the stable `mirai-graph@1.0.0` package. To verify a
+repository checkout directly, use:
 
 ```bash
 git clone https://github.com/zabarov/mirai-graph.git
@@ -53,7 +52,7 @@ npm install
 npm run release:check
 ```
 
-After npm publication, external projects will use:
+External projects use:
 
 ```bash
 npm install -D mirai-graph
@@ -68,6 +67,7 @@ npx mirai-graph detect . --markdown
 npx mirai-graph bootstrap . --mode suggest --markdown
 npx mirai-graph init . --profile software_specification
 npx mirai-graph validate .
+npx mirai-graph migrate .
 ```
 
 What each command does:
@@ -76,16 +76,51 @@ What each command does:
 - `bootstrap --mode suggest` creates proposal/evidence only.
 - `init` creates starter graph files after you choose a profile.
 - `validate` checks the graph package structure and profile rules.
+- `migrate` shows a read-only plan; add `--apply` only after reviewing it.
 
 If you are using a repository checkout before npm publication, run the same
 commands through `node packages/cli/mirai-graph.js` from this repository.
 
+## Project Technology
+
+Project Technology is the executable part of Mirai Graph. It gives ordinary
+projects, skills, platforms and federations one safe way to build task context,
+bind an accepted target and verify that work still follows it.
+
+Enable it in the existing root manifest:
+
+```json
+{
+  "extensions": {
+    "mirai.project_technology": {
+      "contract_version": "1.0.0",
+      "enabled": true,
+      "context_policy": "task_scoped",
+      "source_boundary": "hybrid_sot"
+    }
+  }
+}
+```
+
+Read-only operations never need `--apply`:
+
+```bash
+mirai-graph technology explain .
+mirai-graph technology status .
+mirai-graph technology context . --task "implement the approved feature"
+mirai-graph technology verify . --significant-work
+```
+
+Changing operations return a preview unless `--apply` is explicit. See
+[Project Technology](standard/project-technology.md) for provider/consumer
+binding, local state and safety boundaries.
+
 ## What Files Will Be Created?
 
 For a typical software project, `init` creates:
 
 ```text
-mirai-graph-package.json
+graph.json
 graph/
   objects.json      # features, requirements, risks, decisions, evidence
   relations.json    # depends_on, implements, blocks, evidences, governs
@@ -162,10 +197,13 @@ releases/      Release notes and release process.
 - [Context Pack Generation](standard/context-pack-generation.md)
 - [Governance](standard/governance.md)
 - [Hybrid Source Of Truth](standard/hybrid-source-of-truth.md)
+- [Profile Boundaries](standard/profile-boundaries.md)
 - [Implementation Control](standard/implementation-control.md)
 - [Process Control Contract](standard/process-control-contract.md)
 - [Technology Quality Feedback](standard/technology-quality-feedback.md)
 - [Dynamic Episode Layer](standard/dynamic-episode-layer.md)
+- [Semantic Intent Resolution](standard/semantic-intent-resolution.md)
+- [Goal Vector Quality Control](standard/goal-vector-quality-control.md)
 - [Character Layer Profile](profiles/character-layer/README.md)
 - [Organization Governance](standard/organization-governance.md)
 
@@ -226,13 +264,31 @@ Generated context, evidence, feedback, cockpit metrics and proposals do not
 authorize canonical updates by themselves. Controlled updates require explicit
 governance gates.
 
+Mirai Graph also does not store bulk source content as canonical graph state.
+Documents, code, CMS records, tickets and databases remain source systems. The
+graph stores governable meaning: object ids, relations, summaries, source refs,
+evidence refs, owners, readiness, gates, decisions, episode metadata and Kaizen
+routes. See [Profile Boundaries](standard/profile-boundaries.md).
+
 ## Release Candidate
 
-The current release candidate consolidates the validator-backed public standard
+The stable 1.0 release consolidates the validator-backed public standard
 across the core model, profiles, executable process control, instrumentation,
-Character Layer, Dynamic Episode tracing, adopter kit and evidence boundaries.
+Character Layer, adopter kit and evidence boundaries.
+
+`1.0.0-rc.6` specifically presents the anti-drift / quality-control contour as
+one coherent review surface:
+
+```text
+semantic intent
+-> process and technology control
+-> dynamic episode trace
+-> goal-vector reverse audit
+-> technology quality feedback
+-> Kaizen / replay / regression candidate
+```
 
-See [v1.0.0-rc.4](releases/1.0.0-rc.4.md) and the
+See [v1.0.0-rc.6](releases/1.0.0-rc.6.md) and the
 [Release Process](releases/README.md).
 
 ## What This Repository Is Not

package/ROADMAP.md

@@ -56,6 +56,7 @@ Additional 1.0 RC artifacts now present:
 
 - graph seed and graph embryo;
 - hybrid source of truth;
+- profile boundaries and the extension-before-profile rule;
 - graph-as-AI-brain;
 - action/runtime boundary;
 - implementation-control cycles;
@@ -63,6 +64,8 @@ Additional 1.0 RC artifacts now present:
 - work state machine;
 - process-control contract;
 - technology quality feedback;
+- semantic intent resolution;
+- goal vector quality control;
 - recovery/resume;
 - risk-control matrix;
 - multi-agent coordination;
@@ -73,7 +76,10 @@ Additional 1.0 RC artifacts now present:
 - Character Layer profile, starter pack, integration evidence, portability
   evidence shape and 1.0 readiness artifact.
 - proposal/experimental Dynamic Episode Layer, episode trace schema and
-  synthetic code-generation trace example.
+  synthetic code-generation trace example;
+- `1.0.0-rc.6` anti-drift / quality-control contour that connects semantic
+  intent, process control, dynamic episode tracing, goal-vector reverse audit,
+  technology quality feedback and Kaizen/replay follow-up.
 
 Readiness target:
 
@@ -108,6 +114,8 @@ Post-1.0 gaps:
 - import/export helpers for starter packages;
 - stronger non-software adoption recipes;
 - external adopter feedback from independent pilots.
+- sanitized pilot/replay evidence for the anti-drift contour before claiming
+  real-world drift reduction.
 
 Implemented adopter-workflow focus:
 
@@ -160,6 +168,8 @@ Post-1.0 gaps:
   workflows.
 - evolve dynamic episode traces from synthetic validation to replay/regression
   support for internal pilots.
+- add optional validator/report warnings for profile overlap and bulk-content
+  drift.
 
 Current tooling target:
 
@@ -220,7 +230,7 @@ Planned artifacts:
 
 ## Near-Term Work After 1.0 RC
 
-1. Execute the `1.0.0-rc.4` release: tag, GitHub Release and npm publication
+1. Execute the `1.0.0-rc.6` release: tag, GitHub Release and npm publication
    only after release-state checks are accepted.
 2. Continue adopter workflow: import/export helper skeletons, richer report
    formatting and more adoption recipes when they directly support self-service
@@ -232,4 +242,5 @@ Planned artifacts:
    portability claims.
 6. Explore a local interactive playground only after deterministic report
    fixtures remain stable.
-7. Turn evidence package v0.3 into a publication-ready reproducibility bundle.
+7. Turn evidence package v0.3 and the two article foundations into a
+   publication-ready reproducibility bundle.

package/benchmarks/synthetic-context-reduction-v0/graph.json

@@ -0,0 +1,37 @@
+{
+  "$schema": "https://mirai-graph.dev/schemas/graph-manifest.schema.json",
+  "format": "mirai-graph",
+  "schema_version": "2.0.0",
+  "id": "package.synthetic_context_reduction_v0",
+  "aliases": [],
+  "title": "Synthetic Context-Reduction Benchmark v0",
+  "scope": "package",
+  "kind": "graph_package",
+  "owner": "mirai-graph",
+  "profiles": [
+    "software_specification"
+  ],
+  "graph": {
+    "root": "graph",
+    "source_of_truth": [
+      "graph"
+    ],
+    "objects": [
+      "graph/objects.json"
+    ],
+    "relations": [
+      "graph/relations.json"
+    ],
+    "schemas": [],
+    "generated": [],
+    "raw_sources": [
+      "README.md"
+    ]
+  },
+  "imports": [],
+  "version": "0.1.0",
+  "conformance_level": "level_1",
+  "description": "A public-safe synthetic benchmark for inspecting Mirai Graph context-selection structure.",
+  "license": "CC-BY-4.0",
+  "public_safety": "synthetic_public"
+}

package/docs/README.md

@@ -38,6 +38,7 @@ Core model:
 - [Context Pack Generation](../standard/context-pack-generation.md)
 - [Governance](../standard/governance.md)
 - [Hybrid Source Of Truth](../standard/hybrid-source-of-truth.md)
+- [Profile Boundaries](../standard/profile-boundaries.md)
 
 Executable process and implementation control:
 
@@ -48,6 +49,8 @@ Executable process and implementation control:
 - [Process Control Contract](../standard/process-control-contract.md)
 - [Technology Quality Feedback](../standard/technology-quality-feedback.md)
 - [Dynamic Episode Layer](../standard/dynamic-episode-layer.md)
+- [Semantic Intent Resolution](../standard/semantic-intent-resolution.md)
+- [Goal Vector Quality Control](../standard/goal-vector-quality-control.md)
 - [Episode Trace](../standard/episode-trace.md)
 - [Dynamic Layer For AI-Assisted Code Generation](../standard/dynamic-layer-code-generation.md)
 - [Recovery Resume](../standard/recovery-resume.md)
@@ -92,6 +95,8 @@ Proposal or experimental profile:
 - [Process Transition](../examples/process-transition-minimal/README.md)
 - [Technology Quality Feedback](../examples/technology-quality-feedback-minimal/README.md)
 - [Dynamic Episode Minimal Example](../examples/dynamic-episode-minimal/README.md)
+- [Semantic Intent Resolution](../examples/semantic-intent-resolution-minimal/README.md)
+- [Goal Vector Quality Control](../examples/goal-vector-quality-control-minimal/README.md)
 - [Instrumentation Layer](../examples/instrumentation-layer-minimal/README.md)
 - [Organization Governance](../examples/organization-governance-minimal/README.md)
 - [Societal Governance Proposal Example](../examples/societal-governance-minimal/README.md)
@@ -111,6 +116,8 @@ Proposal or experimental profile:
 - [Roadmap](../ROADMAP.md)
 - [Changelog](../CHANGELOG.md)
 - [Release Process](../releases/README.md)
+- [v1.0.0-rc.6](../releases/1.0.0-rc.6.md)
+- [v1.0.0-rc.5](../releases/1.0.0-rc.5.md)
 - [v1.0.0-rc.4](../releases/1.0.0-rc.4.md)
 - [Contributing](../CONTRIBUTING.md)
 

package/docs/adoption/adopter-kit.md

@@ -114,7 +114,7 @@ npm run validate:adopter-workflow
 
 A 1.0 release-candidate adopter package is ready for first internal use when:
 
-- `mirai-graph-package.json` exists and points to graph, gates and profile.
+- root `graph.json` v2 exists and points to graph sources and profiles.
 - `graph/objects.json` has stable object ids, kinds, summaries and readiness.
 - `graph/relations.json` has directed relations with valid source/target ids.
 - `gates/results.json` records at least the initial validation gate.

package/docs/adoption/choose-profile.md

@@ -10,6 +10,11 @@ the first profile, starter template and validation path.
 Start narrow. A useful first package should model one bounded system, not the
 entire organization.
 
+If a new idea appears to need a new profile, first read
+[Profile Boundaries](../../standard/profile-boundaries.md). Mirai Graph uses
+`extension before profile`: prefer an existing profile object, relation, gate,
+validator or proposal before adding a new profile name.
+
 ## Quick Choice
 
 | You are modeling | Start with profile | Starter template | Validate with |
@@ -96,3 +101,6 @@ node packages/cli/mirai-graph.js adopter report templates/software-project-start
 - Generated context, evidence, feedback and proposals do not authorize
   canonical updates.
 - Production execution stays outside the public alpha adopter workflow.
+- Full documents, articles, code and raw logs should stay in source systems.
+  Mirai Graph stores references, governable meaning, gates and evidence links,
+  not bulk content.

package/docs/adoption/connect-project-15-minutes.md

@@ -17,7 +17,7 @@ From the project you want to connect:
 npm install -D mirai-graph
 ```
 
-Release-candidate note: `mirai-graph@1.0.0-rc.4` is not published on npm yet.
+Release-candidate note: `mirai-graph@1.0.0-rc.6` is not published on npm yet.
 Until npm publication is complete, use a repository checkout and run the same
 commands through `node packages/cli/mirai-graph.js`.
 
@@ -75,7 +75,7 @@ npx mirai-graph init . --profile organization_governance
 `init` creates:
 
 ```text
-mirai-graph-package.json
+graph.json
 graph/objects.json
 graph/relations.json
 gates/results.json

package/docs/adoption/for-developers.md

@@ -71,7 +71,7 @@ package remains structurally usable.
 A starter software graph usually looks like this:
 
 ```text
-mirai-graph-package.json
+graph.json
 graph/
   objects.json
   relations.json

package/docs/adoption/getting-started.md

@@ -25,7 +25,7 @@ npx mirai-graph init . --profile software_specification
 npx mirai-graph validate .
 ```
 
-Release-candidate note: `mirai-graph@1.0.0-rc.4` is not published on npm yet.
+Release-candidate note: `mirai-graph@1.0.0-rc.6` is not published on npm yet.
 Until npm publication is complete, use the repository checkout path and run the
 same commands through `node packages/cli/mirai-graph.js`.
 
@@ -74,7 +74,7 @@ Do not try to model everything at once.
 Recommended minimal structure:
 
 ```text
-mirai-graph-package.json
+graph.json
 graph/
   objects.json
   relations.json

package/docs/adoption/mirai-graph-transition.md

@@ -22,7 +22,8 @@ feedback and controlled evolution.
 
 - Package name: `mirai-graph`.
 - Primary CLI: `mirai-graph`.
-- Preferred package manifest: `mirai-graph-package.json`.
+- Canonical workspace and package manifest: root `graph.json` schema `2.0.0`.
+- `mirai-graph-package.json` and `graph/graph.json` are read-only legacy inputs during the compatibility window.
 - Public docs and examples use Mirai Graph naming.
 
 ## Current Command Surfaces
@@ -33,18 +34,19 @@ The supported current command surfaces are:
 - `node packages/cli/mirai-graph.js ...` for repository checkout maintenance;
 - `node packages/cli/mirai_graph.js ...` as an underscore alias;
 - `node packages/cli/validate-mirai-graph.js ...` for direct validator calls;
-- `mirai-graph-package.json`
+- `graph.json` plus `graph.lock.json` for immutable external pins
 
 The underscore CLI is an alias for local environments that prefer identifier
 style command names. It is not a separate standard.
 
 ## Migration Steps
 
-1. Use `mirai-graph-package.json` for package manifests.
-2. Use `npx mirai-graph validate .` in adopter projects.
-3. Use direct validator calls from repository checkout only when maintaining
+1. Run `mirai-graph migrate .` for a read-only plan and add `--apply` only after review.
+2. Keep exactly one root `graph.json`; do not write a second manifest into `graph/`.
+3. Use `npx mirai-graph validate .` in adopter projects.
+4. Use direct validator calls from repository checkout only when maintaining
    Mirai Graph itself.
-4. Keep generated reports as historical evidence; do not rewrite evidence
+5. Keep generated reports as historical evidence; do not rewrite evidence
    payloads unless they are regenerated by the current tooling.
 
 ## Schema Namespace

package/docs/adoption/seed-to-validated-package.md

@@ -41,7 +41,7 @@ The embryo is a reviewable candidate state. It is not canonical graph state.
 Minimum package shape:
 
 ```text
-mirai-graph-package.json
+graph.json
 graph/
   objects.json
   relations.json

package/docs/repository-purpose.md

@@ -23,7 +23,7 @@ The repository should become the public source of truth for:
 - synthetic benchmarks and playground materials;
 - publications, citations and research evidence.
 
-For the current public release candidate, "source of truth" means public
+For the current stable release, "source of truth" means public
 standardization and reproducible examples. It does not mean that private
 project graphs, internal workflows or generated context packs are published
 here.
@@ -87,7 +87,7 @@ The model is not just a knowledge graph. It combines:
 - implementation and adoption methods.
 - organization-governance coordination patterns.
 
-The current 1.0 release candidate also makes the operating model executable
+The stable 1.0 release also makes the operating model executable
 enough to test:
 
 - launch records define bounded work permission;
@@ -182,7 +182,7 @@ test the model:
 
 Synthetic evidence must be clearly separated from real-world validation.
 
-In the current release candidate, `playground/` is an index of runnable CLI
+In the stable release, `playground/` is an index of runnable CLI
 examples. A hosted or interactive playground is future work.
 
 ### 6. Link Publications And Research Evidence

package/docs/research/evidence-package-v0.3.md

@@ -5,7 +5,7 @@ Status: 1.0 release-candidate public evidence package
 ## Purpose
 
 This package is the current public evidence entrypoint for Mirai Graph
-`1.0.0-rc.4`. It connects three evidence layers:
+`1.0.0-rc.6`. It connects three evidence layers:
 
 1. Synthetic benchmark and baseline comparison.
 2. Executable instrumentation artifacts.
@@ -27,6 +27,9 @@ It does not claim peer-reviewed proof of effectiveness.
 | Instrumentation | `examples/instrumentation-layer-minimal/` | cockpit, traceability and multi-source feedback shape | prediction of real production outcomes |
 | Character Layer | `examples/character-layer-readiness-1-0/` | reusable behavior-governance profile, integration shape and portability evidence format | proof of broad model equivalence or production replacement safety |
 | Dynamic Episode | `examples/dynamic-episode-minimal/` | event-to-decision trace shape, blocked alternatives and feedback routing | proof of reduced drift in real code generation |
+| Semantic Intent | `examples/semantic-intent-resolution-minimal/` | input-control shape for intent, scale, owner, missing context and safe default action | proof that all real requests are interpreted correctly |
+| Goal Vector | `examples/goal-vector-quality-control-minimal/` | final-outcome vector links, reverse audit and correction routing | proof of reduced real-world goal drift |
+| Technology Feedback | `examples/technology-quality-feedback-minimal/` | conformance-check shape and classified process findings | proof that real work always followed the technology |
 | Public pilots | `pilots/independent-implementation-001..005-*` | profile applicability across domains | proof of real-world effectiveness |
 | Adopter kit | `docs/adoption/adopter-kit.md` | repeatable first adoption path | successful adoption without review |
 
@@ -54,6 +57,14 @@ H7. Externalized Character Layer constraints can improve behavior portability
 across model backends when measured through the same graph context, task set and
 fixtures.
 
+H8. Semantic intent resolution plus goal-vector quality control can reduce
+wrong-start errors, goal drift and false completion claims in AI-assisted
+workflows.
+
+H9. Dynamic episode traces plus technology quality feedback can make
+AI-assisted work auditable enough to support replay, regression and Kaizen
+improvement loops.
+
 These are hypotheses and engineering claims under test, not settled scientific
 facts.
 
@@ -67,6 +78,8 @@ facts.
 | Process transition | `examples/process-transition-minimal/` | `npm run validate:process-transition` | Executable fixture |
 | Process explanation report | `examples/process-transition-minimal/results/` | `npm run validate:process-transition-report` | Executable fixture |
 | Technology feedback | `examples/technology-quality-feedback-minimal/` | `npm run validate:technology-quality-feedback` | Executable fixture |
+| Semantic intent | `examples/semantic-intent-resolution-minimal/` | `npm run validate:semantic-intent-resolution` | Executable fixture |
+| Goal vector quality | `examples/goal-vector-quality-control-minimal/` | `npm run validate:goal-vector-quality-control` | Executable fixture |
 | Instrumentation layer | `examples/instrumentation-layer-minimal/` | `npm run validate:instrumentation-layer` | Executable fixture |
 | Instrumentation report | `examples/instrumentation-layer-minimal/` | `npm run validate:instrumentation-report` | Readable report |
 | Character Layer readiness | `examples/character-layer-readiness-1-0/` | `npm run validate:character-layer-readiness` | Profile readiness |
@@ -102,6 +115,8 @@ The public suite currently measures or validates:
 - process-transition validity;
 - missing transition evidence;
 - technology-feedback classification;
+- semantic-intent route decision shape;
+- goal-vector linkage and reverse-audit shape;
 - instrumentation report validity;
 - character-layer readiness;
 - model-portability evidence shape;
@@ -172,7 +187,7 @@ node packages/cli/mirai-graph.js report playground
 
 ## Current Evidence Verdict
 
-Mirai Graph `1.0.0-rc.4` supports these bounded claims:
+Mirai Graph `1.0.0-rc.6` supports these bounded claims:
 
 - a public executable standard draft exists;
 - schemas, validators, templates, examples and pilots are runnable;
@@ -186,13 +201,20 @@ Mirai Graph `1.0.0-rc.4` supports these bounded claims:
 - Dynamic Episode traces can record a selected path, blocked alternatives,
   evidence refs, finding classification and feedback routing for one governed
   episode;
+- Semantic Intent Resolution can represent and validate an intent, scale,
+  owner/process and safe-default-action decision before execution;
+- Goal Vector Quality Control can represent and validate final-outcome vector
+  links, reverse audit, drift findings and correction routes;
+- the current public anti-drift contour can be inspected as a connected
+  sequence of input control, process control, episode trace, goal-vector audit,
+  technology feedback and Kaizen/replay candidate;
 - synthetic baseline comparison can record context, evidence and assumption
   metrics;
 - public-safe pilots cover project-management, software-specification,
   AI-employee, research-program and organization-governance scenarios;
 - the adopter kit gives a reproducible first-use path.
 
-Mirai Graph `1.0.0-rc.4` does not yet support these claims:
+Mirai Graph `1.0.0-rc.6` does not yet support these claims:
 
 - peer-reviewed effectiveness;
 - production autonomous execution safety;
@@ -202,6 +224,8 @@ Mirai Graph `1.0.0-rc.4` does not yet support these claims:
 - automatic correctness of generated context;
 - automatic canonical update authorization from evidence, feedback or
   proposals.
+- proven real-world reduction of wrong-start, goal drift or process violation
+  rates without pilot/replay/regression evidence.
 
 ## Next Evidence Work
 
@@ -214,6 +238,8 @@ Mirai Graph `1.0.0-rc.4` does not yet support these claims:
 - Add reviewer-effort and time-to-relevant-context metrics.
 - Prepare a publication-ready reproducibility bundle that cites this package as
   the current release-candidate evidence index.
+- Use the two article foundation packages in `publications/` to collect
+  sanitized pilot evidence and publication-ready limitations.
 
 ## Public Safety Boundary