mnemon-dev · Grivn · Aug 9, 2026 · Aug 9, 2026 · Aug 9, 2026 · Aug 9, 2026
diff --git a/.github/PULL_REQUEST_TEMPLATE.md b/.github/PULL_REQUEST_TEMPLATE.md
@@ -8,7 +8,8 @@ Link to issue or explain the motivation.
 
 ## Checklist
 
-- [ ] Tests pass (`make test` and `go test ./...`)
+- [ ] Deterministic tests pass (`make test`)
+- [ ] Relevant E2E/process/Docker boundaries pass (`make test-integration`, when affected)
 - [ ] New/changed behavior is covered by tests
 - [ ] Documentation updated (USAGE.md, DESIGN.md, or README) if applicable
 - [ ] User-facing release-note impact described in this PR, if applicable
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -16,6 +16,7 @@ concurrency:
 jobs:
   build-and-test:
     runs-on: ubuntu-latest
+    timeout-minutes: 30
 
     steps:
       - uses: actions/checkout@v4
@@ -31,11 +32,5 @@ jobs:
             go.sum
             harness/go.sum
 
-      - name: Run unit tests
-        run: go test ./...
-
-      - name: Run E2E tests
+      - name: Run deterministic test suite
         run: make test
-
-      - name: Check release and Harness boundary
-        run: make harness-validate
diff --git a/.github/workflows/harness-deep.yml b/.github/workflows/harness-deep.yml
diff --git a/.github/workflows/integration.yml b/.github/workflows/integration.yml
@@ -0,0 +1,33 @@
+name: Manual Integration
+
+on:
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+concurrency:
+  group: integration-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  verify:
+    runs-on: ubuntu-latest
+    timeout-minutes: 60
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          ref: ${{ github.event.pull_request.head.sha || github.sha }}
+
+      - name: Set up Go
+        uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache-dependency-path: |
+            go.sum
+            harness/go.sum
+
+      - name: Run integration suite
+        run: make test-integration
diff --git a/.github/workflows/live.yml b/.github/workflows/live.yml
@@ -0,0 +1,36 @@
+name: Live Agent Evaluation
+
+on:
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+concurrency:
+  group: live-agent-evaluation
+  cancel-in-progress: false
+
+jobs:
+  live:
+    runs-on: ubuntu-latest
+    timeout-minutes: 90
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Go
+        uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache-dependency-path: |
+            go.sum
+            harness/go.sum
+
+      - name: Set up Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: 22.19.0
+
+      - name: Run Pi and DeepSeek live evaluation
+        env:
+          DEEPSEEK_API_KEY: ${{ secrets.DEEPSEEK_API_KEY }}
+        run: make test-live
diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
@@ -23,10 +23,7 @@ jobs:
         with:
           go-version-file: go.mod
 
-      - name: Run unit tests
-        run: go test ./...
-
-      - name: Run E2E tests
+      - name: Run deterministic test suite
         run: make test
 
   release:

diff --git a/AGENTS.md b/AGENTS.md
@@ -3,11 +3,12 @@
 ## Development
 
 - Build with `go build -o mnemon .`.
-- Run the E2E suite with `bash scripts/e2e_test.sh` or `make test`.
-- Validate harness module manifests with `make harness-validate` when changing
-  harness module assets.
-- Run `make harness-quality` for the pinned Harness quality ratchet and
-  `make harness-verify` for the complete local Harness build and test gate.
+- Run `make test` for the deterministic CI suite. It excludes real daemon
+  readiness, CLI E2E, wall-clock scenarios, Docker, and provider calls.
+- Run `make test-integration` explicitly for CLI E2E plus Harness timing, race,
+  process, and Docker boundaries; it is not a regular CI gate.
+- Run `make test-live` only when explicitly validating the paid Pi/DeepSeek
+  scenarios.
 - Treat `harness/` as an experimental, not-yet-released harness layer. Do not
   use it as an implementation dependency for release-path commands such as
   `mnemon setup`; formal integrations belong under `cmd/` and `internal/`.
@@ -25,9 +26,8 @@
 - Keep authority, digest, fence, bounds, CAS cardinality, and fail-closed checks
   explicit. Every goroutine must have an owner, cancellation, bounded work, and
   a wait path.
-- In a scope with a tracked engineering-quality baseline, new or modified code
-  must not increase it. Preserve independent test oracles while compressing
-  fixtures.
+- Preserve independent replay, crash, authorization, and race oracles while
+  compressing fixtures and shared setup.
 
 ## Commit Discipline
 

diff --git a/CLAUDE.md b/CLAUDE.md
@@ -4,6 +4,7 @@
 
 - **Build**: `go build -o mnemon .`
 - **Install**: `make install && mnemon setup`
-- **Test**: `bash scripts/e2e_test.sh`
+- **Test**: `make test`
+- **Integration**: `make test-integration` for CLI E2E and Harness boundaries
 - **Dependencies**: `modernc.org/sqlite`, `spf13/cobra`, `google/uuid`
 - **Optional**: Ollama with `nomic-embed-text` for embedding support
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
@@ -24,15 +24,14 @@ make build
 ## Running Tests
 
 ```bash
-make unit       # Go unit tests (go test ./...)
-make test       # Full E2E test suite (scripts/e2e_test.sh)
-make vet        # Static analysis (go vet ./...)
-make harness-quality # Pinned Harness format, static, architecture, and debt ratchets
-make harness-verify  # Harness binaries, declarations, quality, vet, and unit tests
+make test              # Required deterministic CI suite for both Go modules
+make test-integration  # Opt-in CLI E2E, timing, race, process, and Docker suite
+make test-live         # Explicit paid Pi/DeepSeek evaluation
 ```
 
-Both `make unit` and `make test` must pass before submitting a PR. A change to
-the experimental Harness must also pass `make harness-verify`.
+`make test` must pass before submitting a PR. Run `make test-integration`
+proportionally when changing CLI E2E behavior or Harness process, timing,
+transport, or Docker boundaries; it is intentionally outside regular CI.
 
 ## Code Style
 
@@ -41,8 +40,7 @@ the experimental Harness must also pass `make harness-verify`.
 - All exported functions and types must have doc comments
 - Use `fmt.Errorf("context: %w", err)` for error wrapping
 
-For architecture, concurrency, persistence, abstraction, test design, and the
-incremental quality ratchet, follow the
+For architecture, concurrency, persistence, abstraction, and test design, follow the
 [Go Engineering Standard](docs/development/go-engineering-standard.md). Design
 patterns and Go language features are tools for reducing change amplification,
 not quotas or substitutes for explicit safety checks.
@@ -64,8 +62,8 @@ imperative form. The CHANGELOG filter excludes `docs:`, `test:`, `ci:`, and
 ## Submitting Changes
 
 1. Fork the repository and create a feature branch from `master`.
-2. Make your changes and run the gates that cover them, including
-   `make harness-verify` for Harness changes.
+2. Make your changes and run the proportional test level, including
+   `make test-integration` for affected Harness boundary behavior.
 3. Update documentation (USAGE.md, DESIGN.md, or README) if your change affects user-facing behavior.
 4. For user-facing changes, describe the release-note impact in the PR body. Maintainers update `CHANGELOG.md` during release preparation unless they explicitly ask for a changelog entry in the PR.
 5. Open a pull request against `master`.

diff --git a/Makefile b/Makefile
@@ -9,14 +9,30 @@ HARNESS_LDFLAGS := -s -w -X main.version=$(VERSION)
 GO_VERSION   := $(shell awk '$$1 == "go" { print $$2; exit }' go.mod)
 HARNESS_GO_VERSION := $(shell awk '$$1 == "go" { print $$2; exit }' harness/go.mod)
 HARNESS_GO   := env GOTOOLCHAIN=go$(HARNESS_GO_VERSION) GOFLAGS=-mod=readonly go -C harness
+HARNESS_DETERMINISTIC_PKGS := \
+	./cmd/mnemon-harness \
+	./internal/agency \
+	./internal/attach \
+	./internal/authority \
+	./internal/cas \
+	./internal/selector \
+	./internal/selector/simtest \
+	./test/architecture \
+	./test/observer \
+	./test/r7/domainops/trace \
+	./test/r8/network/runner/trace
+HARNESS_TESTDATA_PKGS := \
+	./internal/selector/testdata/network/cmd/r8-peer \
+	./testdata/r7/domain-ops/cmd/domain-load \
+	./testdata/r7/domain-ops/cmd/domain-world \
+	./testdata/r7/domain-ops/cmd/domainctl \
+	./testdata/r7/domain-ops/world
 GOBIN       := $(shell go env GOBIN)
 ifeq ($(GOBIN),)
   GOBIN     := $(shell go env GOPATH)/bin
 endif
 
-.PHONY: deps build harness-build install uninstall test unit vet harness-validate harness-quality harness-verify
-.PHONY: harness-live-pi
-.PHONY: harness-r8 harness-r8-docker harness-domain-ops harness-domain-ops-live
+.PHONY: deps build harness-build install uninstall test test-integration test-live
 .PHONY: docker-build docker-run compose-up compose-down compose-dev release-snapshot clean help
 
 .DEFAULT_GOAL := help
@@ -48,46 +64,25 @@ uninstall: ## Remove mnemon binary from $GOBIN
 
 # ── Test ─────────────────────────────────────────────────────────────
 
-test: ## Run E2E test suite (the script builds the tested binary once)
-	bash scripts/e2e_test.sh
-
-unit: ## Run Go unit tests
-	go test ./...
-
-vet: ## Run go vet static analysis
+test: ## Run deterministic tests without E2E, real daemon, or provider calls
 	go vet ./...
+	go test ./...
+	$(HARNESS_GO) vet ./... $(HARNESS_TESTDATA_PKGS)
+	$(HARNESS_GO) test $(HARNESS_DETERMINISTIC_PKGS) -count=1
 
-harness-validate: ## Validate the R7 projection, contract, and evidence bindings
-	$(HARNESS_GO) test ./internal/attach ./tools/corecontract ./test/contracts -count=1
-
-harness-quality: ## Run pinned, non-mutating Harness quality gates
-	@base_ref="$${HARNESS_QUALITY_BASE_REF:-HEAD}"; \
-		$(HARNESS_GO) run ./tools/quality check --root .. --base-ref "$$base_ref"
-	$(HARNESS_GO) vet ./...
-	$(HARNESS_GO) test ./tools/quality -count=1
-
-harness-live-pi: ## Run the opt-in Pi/DeepSeek live smoke
-	@test "$${LIVE_PI:-}" = 1 || { echo "error: set LIVE_PI=1" >&2; exit 2; }
-	@test -n "$${DEEPSEEK_API_KEY:-}" || { echo "error: DEEPSEEK_API_KEY is required" >&2; exit 2; }
-	harness/test/r7/runner/run_live_pi.sh
-
-harness-r8: ## Test the optional, removable R8 selector and its proof adapters
-	$(HARNESS_GO) test ./internal/selector ./internal/selector/simtest ./internal/selector/testdata/network/cmd/r8-peer -count=1
-	$(HARNESS_GO) test -race ./internal/selector ./internal/selector/simtest ./internal/selector/testdata/network/cmd/r8-peer -count=1
-
-harness-r8-docker: harness-r8 ## Run the isolated five-peer R8 network proof
-	harness/test/r8/network/runner/run_docker.sh
-
-harness-domain-ops: ## Run the opt-in real-service federated operations world
+test-integration: ## Run opt-in E2E, timing, race, process, and Docker tests
+	bash scripts/e2e_test.sh
+	$(HARNESS_GO) test -p 1 ./... $(HARNESS_TESTDATA_PKGS) -count=1
+	$(HARNESS_GO) test -race -p 1 ./internal/... $(HARNESS_TESTDATA_PKGS) -count=1
+	harness/test/r7/runner/run_cases.sh
+	harness/test/r7/runtime/pi/run_delegate_oracle.sh
 	harness/test/r7/domainops/run_world.sh
+	harness/test/r8/network/runner/run_docker.sh
 
-harness-domain-ops-live: ## Run the paid autonomous Pi/DeepSeek operations case
-	@test "$${LIVE_DOMAIN_OPS:-}" = 1 || { echo "error: set LIVE_DOMAIN_OPS=1" >&2; exit 2; }
+test-live: ## Run the paid Pi/DeepSeek smoke and federated operations case
 	@test -n "$${DEEPSEEK_API_KEY:-}" || { echo "error: DEEPSEEK_API_KEY is required" >&2; exit 2; }
-	harness/test/r7/domainops/run_live.sh
-
-harness-verify: harness-quality ## Run the complete exact-tree R7 evidence gate and write its report
-	$(HARNESS_GO) run ./tools/corecontract/cmd/core-gate --root ..
+	LIVE_PI=1 harness/test/r7/runner/run_live_pi.sh
+	LIVE_DOMAIN_OPS=1 harness/test/r7/domainops/run_live.sh
 
 # ── Containers / Deployment ──────────────────────────────────────────
 

diff --git a/README.md b/README.md
@@ -382,7 +382,8 @@ Mnemon architecture.
 ```bash
 make build          # build binary
 make install        # build + install to $GOBIN
-make test           # run E2E test suite
+make test           # run deterministic CI tests
+make test-integration  # opt-in CLI E2E and Harness boundary tests
 mnemon setup        # interactive setup
 mnemon setup --eject  # remove all integrations
 make help           # show all targets
@@ -395,7 +396,7 @@ See [Development and Deployment](docs/DEPLOYMENT.md) for Docker, Compose, Ollama
 ## Documentation
 
 - [Mnemon Harness Beta](harness/README.md) — experimental host-agent lifecycle state
-- [Go Engineering Standard](docs/development/go-engineering-standard.md) — maintainability, concurrency, persistence, testing, and quality ratchets
+- [Go Engineering Standard](docs/development/go-engineering-standard.md) — maintainability, concurrency, persistence, testing, and review thresholds
 - [Design & Architecture](docs/DESIGN.md) — current engine architecture, algorithms, integration design
 - [Usage & Reference](docs/USAGE.md) — CLI commands, embedding support, architecture overview
 - [Memory Import Guide](docs/IMPORT.md) — schema and LLM prompt for importing historical chats

diff --git a/docs/DEPLOYMENT.md b/docs/DEPLOYMENT.md
@@ -6,15 +6,15 @@ Prerequisites:
 
 - Go 1.24.6 or newer in the 1.24 series
 - `make`
-- `jq` for the E2E test script
+- `jq` only when running the opt-in CLI E2E/integration suite
 
 Common commands:
 
 ```bash
 make deps
 make build
-make unit
-make test
+make test              # deterministic CI suite
+make test-integration  # optional E2E/process/Docker suite
 make install
 ```