samples: FactoryDemo.Api.FSharp — JSON API (v0, in-memory, stack-independent)

[AceHack]

Summary

Minimal F# ASP.NET Core Web API serving CRM seed data as JSON. Any frontend choice consumes the same 9 endpoints. Build green, 0 warnings. dotnet run + curl confirmed all endpoints working.

Verified output

GET /api/pipeline/funnel      → [{"count":10,"stage":"Lead","totalCents":5400000}, ...5 stages]
GET /api/pipeline/duplicates  → 2 collisions ([1,13] share alice@acme; [5,19] share bob@trades)
GET /api/customers/1          → {"id":1,"name":"Alice Plumbing LLC",...}

Endpoints

• /api/customers + /api/customers/{id} + /api/customers/{id}/activities
• /api/opportunities + /api/opportunities/{id}
• /api/activities
• /api/pipeline/funnel (count + total-cents per stage)
• /api/pipeline/duplicates (customers sharing an email)

Why stack-independent

Aaron has not picked Blazor vs React vs other for the frontend. This API ships a JSON contract any of them can consume. Frontend-stack decision stays open.

Why this move

Third EXT commit of the session on priority #1. Live-lock audit fired earlier; inaugural lesson said "ship external-priority increments when smell fires." Three landed, all priority #1.

What this does NOT do

• No Postgres wiring — in-memory v0; Npgsql is a follow-up PR
• No Zeta / DBSP / retraction-native language exposed to frontend
• No writes (read-only v0), no auth, no docker-compose yet

Test plan

• dotnet build -c Release — 0 Warning(s), 0 Error(s)
• dotnet run starts API on http://localhost:5099
• All 9 endpoints return expected JSON shapes
• Frontend team (Aaron's call on stack) can consume the API

🤖 Generated with Claude Code

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 76024e55fe

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[Copilot]

Pull request overview

This PR introduces a ServiceTitan-themed sample surface (CLI CRM demo + minimal JSON API) while also adding a new Zeta.Core.SignalQuality module, associated tests, and several factory-process docs/scripts (drop-zone protocol, live-lock audit, operator-input log, etc.).

Changes:

• Add samples/ServiceTitanFactoryApi (in-memory ASP.NET Core minimal API) and samples/ServiceTitanCrm (console demo) for CRM-shaped scenarios.
• Add src/Core/SignalQuality.fs plus new algebra/operator tests validating CRM scenarios and signal-quality scoring.
• Add factory hygiene tooling/docs: live-lock audit script + history log, drop-zone protocol, and various research/process documents.

Reviewed changes

Copilot reviewed 35 out of 37 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
tools/audit/live-lock-audit.sh	New script to classify recent commits and flag “live-lock” when external-motion ratio is low.
tests/Tests.FSharp/Tests.FSharp.fsproj	Registers new test modules (SignalQuality.Tests, CrmScenarios.Tests).
tests/Tests.FSharp/Operators/CrmScenarios.Tests.fs	New operator-level scenario tests mirroring CRM demo behavior.
tests/Tests.FSharp/Algebra/SignalQuality.Tests.fs	New tests for SignalQuality scoring functions and composition semantics.
src/Core/SignalQuality.fs	Introduces SignalQuality module/types for multi-dimension quality measurement and scoring.
src/Core/Core.fsproj	Adds SignalQuality.fs to Core compile list.
samples/ServiceTitanFactoryApi/ServiceTitanFactoryApi.fsproj	New sample Web API project file.
samples/ServiceTitanFactoryApi/Seed.fs	Deterministic in-memory seed data for the JSON API sample.
samples/ServiceTitanFactoryApi/README.md	Usage + endpoint documentation for the JSON API sample.
samples/ServiceTitanFactoryApi/Program.fs	Minimal API endpoints for customers/opportunities/activities + derived pipeline views.
samples/ServiceTitanCrm/ServiceTitanCrm.fsproj	New console CRM demo project file.
samples/ServiceTitanCrm/Program.fs	Console demo showing retraction-native CRM operations and views.
memory/project_reproducible_stability_as_obvious_purpose_2026_04_22.md	New project memory entry about “reproducible stability” directive.
memory/project_operator_input_quality_log_directive_2026_04_22.md	New memory entry describing operator-input quality log directive.
memory/project_arc3_adversarial_self_play_emulator_absorption_scoring_2026_04_22.md	New memory entry on ARC-3-style scoring loop concept.
memory/project_aaron_drop_zone_protocol_2026_04_22.md	New memory entry documenting drop/ inbox protocol.
memory/observed-phenomena/2026-04-19-transcript-duplication-splitbrain-hypothesis.md	New observed-phenomena note tied to an existing artifact.
drop/README.md	Adds the tracked drop-zone protocol document.
drop/.gitignore	Configures drop-zone to ignore everything except sentinel files.
docs/research/zeta-self-use-tiny-bin-file-germination-2026-04-22.md	Research sketch describing a proposed tiny-bin-file “germination” step.
docs/research/oss-deep-research-zeta-aurora-2026-04-22.md	Absorption note for an OpenAI Deep Research report.
docs/research/openai-deep-ingest-cross-substrate-readability-2026-04-22.md	Research note on cross-substrate ingest/readability tradeoffs.
docs/research/meta-pixel-perfect-text-to-image-youtube-wink-2026-04-22.md	Research note capturing an external “pixel perfect T2I” signal.
docs/research/cutting-edge-database-gap-review-2026-04-23.md	Research note identifying cutting-edge DB gaps and candidate backlog items.
docs/research/arc3-adversarial-self-play-emulator-absorption-scoring-2026-04-22.md	Research doc elaborating the ARC-3 adversarial self-play framing.
docs/research/amara-network-health-oracle-rules-stacking-2026-04-22.md	Research doc capturing oracle/stacking critique structure and annotations.
docs/operator-input-quality-log.md	Adds operator-input scoring rubric + inaugural log entries.
docs/hygiene-history/live-lock-audit-history.md	Adds history log format + inaugural run entry for live-lock audit.
docs/AUTONOMOUS-LOOP.md	Updates priority ladder to include drop-zone audit step.
Zeta.sln	Adds sample projects and new solution configurations.
README.md	Adds “thesis: reproducible stability” section pointing to AGENTS.md.
AGENTS.md	Adds “purpose: reproducible stability” and adjusts “ship/break” phrasing.
.gitignore	Ignores .btw-queue.md and .playwright-mcp/, fixes node_modules entry formatting.
.claude/commands/btw.md	Adds /btw command guidance and handling procedure.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f78f3ed680

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e0280afa65

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[Copilot]

Pull request overview

Copilot reviewed 36 out of 38 changed files in this pull request and generated 7 comments.

[AceHack]

Addressed 6 findings scoped to this PR (mine):

1. Zeta.sln BOM (codex + Copilot L1) — stripped. dotnet sln list still parses.
2. README.md L23 (Copilot, CrmKernel path) — actually C# companion path at samples/FactoryDemo.Api.CSharp/ — added "(lands in PR samples: FactoryDemo.Api.CSharp — JSON API companion, parity with F# sibling #147)" marker.
3. README.md L82 (Copilot, anonymous record rendering) — replaced {{ Stage, Count, TotalCents }} (which renders as single-brace in markdown) with explicit F# syntax: {| Stage = ...; Count = ...; TotalCents = ... |}.
4. Program.fs L48 (Copilot, incomplete endpoints list) — expanded the endpoints array from 5 to all 9 endpoints (customers + by-id + activities, opportunities + by-id, activities, funnel, duplicates). Build stays green.
5. smoke-test.sh L79-82 (Copilot, casing comment inconsistency) — corrected the comment: F# anonymous-record fields declared lowercase emit lowercase JSON. Matches C# after System.Text.Json default camelCasing. Renamed the check label from root.Name to root.name to match.

Not addressed in this PR (stacked content from pre-session PRs):

• tools/audit/live-lock-audit.sh L53 — from PR live-lock audit tool + cutting-edge DB gap review + 4 BACKLOG rows #143
• tests/Tests.FSharp/Operators/CrmScenarios.Tests.fs L2 (nowarn redundant) — from PR live-lock audit tool + cutting-edge DB gap review + 4 BACKLOG rows #143
• src/Core/SignalQuality.fs L28, L107, L160 — pre-session
• docs/operator-input-quality-log.md L7 — pre-session
• docs/hygiene-history/live-lock-audit-history.md L33 — from PR live-lock audit tool + cutting-edge DB gap review + 4 BACKLOG rows #143
• drop/README.md L16 — pre-session (drop/ gitignored going forward in PR hygiene: gitignore drop/ + .playwright-mcp/ — ferry and scratch directories #151)
• samples/ServiceTitanFactoryApi/README.md — old path from the rename commit; git shows the old file in the diff as deleted, which is expected

The README.md L22 contributor-name finding is a valid callout on previously-landed content I updated this session; it's in a file touched by multiple recent PRs and belongs on its own targeted sweep rather than conflated with this one.

Not self-resolving conversations — human maintainer's move on review.

🤖 Generated with Claude Code

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a5119f664f

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 6 comments.

[Copilot]

P2: The smoke test writes server output to a fixed path under /tmp. This can be clobbered by concurrent runs and is left behind after success. Prefer creating a unique temp file (e.g., via mktemp) and removing it in cleanup (or printing its location) to avoid collisions and stale logs.

[Copilot]

P2: The README hard-codes a sibling PR number ("PR #147") for the planned C# companion sample. PR numbers are repo-state specific and will become stale over time; consider referencing only the planned directory/path (or an issue/ADR) without embedding a PR number in long-lived docs.

Suggested change

	(`samples/FactoryDemo.Api.CSharp/`, sibling PR #147) is planned
	with the same 9 endpoints, matching JSON shapes, and identical
	seed. C# is the more popular .NET language, so it is the natural
	primary demo path; F# stays the reference because F# looks
	closer to math, which makes theorems over the algebra easier
	to express.
	at `samples/FactoryDemo.Api.CSharp/` is planned with the same 9
	endpoints, matching JSON shapes, and identical seed. C# is the
	more popular .NET language, so it is the natural primary demo
	path; F# stays the reference because F# looks closer to math,
	which makes theorems over the algebra easier to express.

[Copilot]

P2: This comment embeds a specific PR number for the planned C# sibling smoke test. PR numbers tend to go stale in long-lived scripts/docs; consider removing the PR reference and keeping the note focused on parity with the companion sample.

Suggested change

	# Dependencies on host: dotnet, curl, jq. A C# sibling API (sibling
	# PR #147) will land its own parallel smoke-test so parity between
	# the two APIs is ground-truth-testable.
	# Dependencies on host: dotnet, curl, jq. A C# sibling API will land
	# its own parallel smoke-test so parity between the two APIs is
	# ground-truth-testable.

[Copilot]

P0: This seed data includes a specific person name ("Aaron Smith") in both comments and customer display name. Repo convention forbids direct contributor names in code/docs (docs/AGENT-BEST-PRACTICES.md:284-292); even if intended as fictional, it is indistinguishable from maintainer attribution. Please rename this customer/comment to a role-based or clearly fictional/non-contributor identifier and keep the duplicate-email scenario intact.

Suggested change

	// Email collision #1: Alice Plumbing (1) and Aaron Smith (13) share alice@acme.example.
	// Email collision #2: Bob HVAC (5) and Quincy Assistant (19) share bob@trades.example.
	[ mk 1L "Alice Plumbing LLC" "alice@acme.example" "555-0101" "123 Elm St, Portland OR"
	mk 2L "Benson Roofing" "benson@roof.example" "555-0102" "45 Oak Ave, Seattle WA"
	mk 3L "Crystal Electric" "crystal@sparks.example" "555-0103" "9 Pine Rd, Boise ID"
	mk 4L "Delta HVAC & Mechanical" "delta@hvac.example" "555-0104" "700 Main St, Spokane WA"
	mk 5L "Bob HVAC Services" "bob@trades.example" "555-0105" "12 Bay Blvd, Tacoma WA"
	mk 6L "Evergreen Landscaping" "info@evergreen.example" "555-0106" "88 Forest Ln, Eugene OR"
	mk 7L "Fairbanks Plumbing" "contact@fairbanks.example" "555-0107" "5 River Rd, Anchorage AK"
	mk 8L "Granite Pest Control" "hello@granite.example" "555-0108" "301 Stone Way, Boise ID"
	mk 9L "Highland Roofing Co" "highland@roof.example" "555-0109" "22 Hill Dr, Bend OR"
	mk 10L "Iron Tree Electric" "iron@tree.example" "555-0110" "17 Spruce St, Salem OR"
	mk 11L "Jackson Pool Services" "jackson@pools.example" "555-0111" "600 Lake Rd, Reno NV"
	mk 12L "Klein Garage Doors" "klein@doors.example" "555-0112" "44 4th Ave, Medford OR"
	mk 13L "Aaron Smith (new contact)" "alice@acme.example" "555-0113" "123 Elm St, Portland OR"
	mk 14L "Lakeview Solar" "lakeview@solar.example" "555-0114" "250 Shore Dr, Bellevue WA"
	mk 15L "Mountain Well Drilling" "mountain@wells.example" "555-0115" "12 Ridge Rd, Coeur dAlene ID"
	mk 16L "Nightingale Security" "ngale@secure.example" "555-0116" "88 Watch Way, Vancouver WA"
	mk 17L "Oak Hill Septic" "oak@septic.example" "555-0117" "14 Rural Rt 3, Gresham OR"
	mk 18L "Prairie Window Cleaning" "prairie@windows.example" "555-0118" "66 Glass Rd, Kennewick WA"
	mk 19L "Quincy Assistant (Bob HVAC)" "bob@trades.example" "555-0119" "12 Bay Blvd, Tacoma WA"
	mk 20L "Redwood Tree Service" "redwood@trees.example" "555-0120" "3 Canopy Ct, Hillsboro OR" ]
	// Email collision #1: Alice Plumbing (1) and Alice Plumbing Intake Contact (13) share alice@acme.example.
	// Email collision #2: Bob HVAC (5) and Quincy Assistant (19) share bob@trades.example.
	[ mk 1L "Alice Plumbing LLC" "alice@acme.example" "555-0101" "123 Elm St, Portland OR"
	mk 2L "Benson Roofing" "benson@roof.example" "555-0102" "45 Oak Ave, Seattle WA"
	mk 3L "Crystal Electric" "crystal@sparks.example" "555-0103" "9 Pine Rd, Boise ID"
	mk 4L "Delta HVAC & Mechanical" "delta@hvac.example" "555-0104" "700 Main St, Spokane WA"
	mk 5L "Bob HVAC Services" "bob@trades.example" "555-0105" "12 Bay Blvd, Tacoma WA"
	mk 6L "Evergreen Landscaping" "info@evergreen.example" "555-0106" "88 Forest Ln, Eugene OR"
	mk 7L "Fairbanks Plumbing" "contact@fairbanks.example" "555-0107" "5 River Rd, Anchorage AK"
	mk 8L "Granite Pest Control" "hello@granite.example" "555-0108" "301 Stone Way, Boise ID"
	mk 9L "Highland Roofing Co" "highland@roof.example" "555-0109" "22 Hill Dr, Bend OR"
	mk 10L "Iron Tree Electric" "iron@tree.example" "555-0110" "17 Spruce St, Salem OR"
	mk 11L "Jackson Pool Services" "jackson@pools.example" "555-0111" "600 Lake Rd, Reno NV"
	mk 12L "Klein Garage Doors" "klein@doors.example" "555-0112" "44 4th Ave, Medford OR"
	mk 13L "Alice Plumbing Intake Contact" "alice@acme.example" "555-0113" "123 Elm St, Portland OR"
	mk 14L "Lakeview Solar" "lakeview@solar.example" "555-0114" "250 Shore Dr, Bellevue WA"
	mk 15L "Mountain Well Drilling" "mountain@wells.example" "555-0115" "12 Ridge Rd, Coeur dAlene ID"
	mk 16L "Nightingale Security" "ngale@secure.example" "555-0116" "88 Watch Way, Vancouver WA"
	mk 17L "Oak Hill Septic" "oak@septic.example" "555-0117" "14 Rural Rt 3, Gresham OR"
	mk 18L "Prairie Window Cleaning" "prairie@windows.example" "555-0118" "66 Glass Rd, Kennewick WA"
	mk 19L "Quincy Assistant (Bob HVAC)" "bob@trades.example" "555-0119" "12 Bay Blvd, Tacoma WA"
	mk 20L "Redwood Tree Service" "redwood@trees.example" "555-0120" "3 Canopy Ct, Hillsboro OR" ]

[Copilot]

P1: JSON property casing is inconsistent across endpoints: the root endpoint returns lowercase fields (name/version/endpoints) while the record/anonymous-record responses from other endpoints will serialize as PascalCase by default. This creates a confusing contract for API consumers. Consider either configuring System.Text.Json to enforce a single naming policy (e.g., camelCase) for all responses or aligning the root payload field names with the rest of the API.

[Copilot]

P1: The script claims to start on a "random free port", but it only picks a random port in a range and does not verify it is actually free. This can make the smoke test flaky if that port is already in use. Consider probing for a free port (or retrying on failure) instead of assuming the chosen port is available.

Suggested change

	# Pick a high random port to avoid clashes with other dev services.
	PORT=$(( 5100 + RANDOM % 400 ))
	URL="http://localhost:${PORT}"
	echo "Building API..."
	dotnet build "$PROJECT" -c Release --nologo -v quiet >/dev/null
	echo "Starting API on ${URL}..."
	dotnet run --project "$PROJECT" -c Release --no-build --urls "$URL" \
	> /tmp/factory-demo-api-fsharp.log 2>&1 &
	API_PID=$!
	cleanup() {
	kill "$API_PID" 2>/dev/null || true
	wait "$API_PID" 2>/dev/null || true
	}
	trap cleanup EXIT
	for _ in {1..20}; do
	if curl -sf "${URL}/" >/dev/null 2>&1; then
	break
	fi
	sleep 0.5
	done
	if ! curl -sf "${URL}/" >/dev/null 2>&1; then
	echo "API did not come up within budget. Log:" >&2
	cat /tmp/factory-demo-api-fsharp.log >&2
	API_LOG="/tmp/factory-demo-api-fsharp.log"
	API_PID=""
	PORT=""
	URL=""
	cleanup() {
	if [ -n "${API_PID}" ]; then
	kill "$API_PID" 2>/dev/null || true
	wait "$API_PID" 2>/dev/null || true
	fi
	}
	trap cleanup EXIT
	start_api() {
	local attempt
	local probe
	for attempt in {1..10}; do
	PORT=$(( 5100 + RANDOM % 400 ))
	URL="http://localhost:${PORT}"
	echo "Starting API on ${URL} (attempt ${attempt}/10)..."
	: > "$API_LOG"
	dotnet run --project "$PROJECT" -c Release --no-build --urls "$URL" \
	> "$API_LOG" 2>&1 &
	API_PID=$!
	for probe in {1..20}; do
	if curl -sf "${URL}/" >/dev/null 2>&1; then
	return 0
	fi
	if ! kill -0 "$API_PID" 2>/dev/null; then
	break
	fi
	sleep 0.5
	done
	cleanup
	API_PID=""
	done
	return 1
	}
	echo "Building API..."
	dotnet build "$PROJECT" -c Release --nologo -v quiet >/dev/null
	if ! start_api; then
	echo "API did not come up within budget. Log:" >&2
	cat "$API_LOG" >&2