Add current_tag attribute to post_index component

Production build broke because our custom lint rule was compiled. The
credo linter is not available and not necessary in production.

Solution: create separate directory for dev tools.

.devcontainer/Dockerfile

@@ -1,6 +1,11 @@
 FROM mcr.microsoft.com/devcontainers/base:ubuntu-24.04
 
 USER root
+RUN apt-get update && apt-get install -y locales && \
+    sed -i '/en_US.UTF-8/s/^# //g' /etc/locale.gen && \
+    locale-gen en_US.UTF-8
+ENV LANG=en_US.UTF-8 LC_ALL=en_US.UTF-8
+
 RUN curl -fsSL https://deb.nodesource.com/setup_lts.x | bash - && \
     apt-get install -y nodejs
 
@@ -19,4 +24,7 @@ RUN /home/vscode/.local/bin/mise exec -- mix local.hex --force && \
 
 RUN npm install -g @mariozechner/pi-coding-agent
 
-USER root
+RUN npm install -g @anthropic-ai/claude-code
+
+RUN echo 'eval "$(/home/vscode/.local/bin/mise activate bash)"' >> /home/vscode/.bashrc
+RUN /home/vscode/.local/bin/mise settings set trusted_config_paths /workspaces/firehose

.devcontainer/devcontainer.json

@@ -1,10 +1,12 @@
 {
     "$schema": "https://containers.dev/implementors/json_schema/",
-    "build": {
-        "dockerfile": "Dockerfile"
-    },
+    "dockerComposeFile": "docker-compose.yml",
+    "service": "app",
+    "workspaceFolder": "/workspaces/firehose",
     "remoteUser": "vscode",
-    "runArgs": [],
+    "containerEnv": {
+        "DB_HOST": "db"
+    },
     "features": {
         "ghcr.io/devcontainers/features/python:1": {},
         "ghcr.io/jsburckhardt/devcontainer-features/uv:1": {},
@@ -17,7 +19,7 @@
         "source=${localEnv:HOME}/.pi/agent/bin,target=/home/vscode/.pi/agent/bin,type=bind,consistency=cached"
     ],
     "postCreateCommand": {
-        "pi-subagents": "bash -ic 'pi install npm:pi-subagents'"
+        "pi-subagents-disabled": "echo 'pi-subagents disabled: upstream JSON schema bug — investigate version pinning separately'"
     },
     "customizations": {
         "jetbrains": {

.devcontainer/docker-compose.yml

@@ -0,0 +1,30 @@
+services:
+  app:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    volumes:
+      - ..:/workspaces/firehose:cached
+    ports:
+      - "4050:4050"
+    command: sleep infinity
+    depends_on:
+      db:
+        condition: service_healthy
+
+  db:
+    image: postgres:16
+    restart: unless-stopped
+    environment:
+      POSTGRES_USER: postgres
+      POSTGRES_PASSWORD: postgres
+    volumes:
+      - pgdata:/var/lib/postgresql/data
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U postgres"]
+      interval: 5s
+      timeout: 5s
+      retries: 5
+
+volumes:
+  pgdata:

.dockerignore

@@ -0,0 +1,31 @@
+# Git
+.git
+
+# Build artifacts
+app/_build
+app/deps
+blogex/_build
+blogex/deps
+
+# Dev/test only
+app/test
+blogex/test
+app/.formatter.exs
+blogex/.formatter.exs
+
+# IDE
+.devcontainer
+.claude
+
+# Documentation
+*.md
+!app/README.md
+
+# Misc
+app/tmp
+app/cover
+app/doc
+blogex/doc
+
+# Dokku setup (may contain secrets)
+dokku-setup.sh

.gitignore

@@ -0,0 +1,3 @@
+# Dokku setup (may contain secrets)
+dokku-setup.sh
+/output/

.pi/skills/test-writer/SKILL.md

@@ -0,0 +1,148 @@
+---
+name: test-writer
+description: Writes tests following Elixir/Phoenix best practices. Ensures DRY tests with proper helper functions, no duplicated setup code, and correct parameter defaults. Use when writing or modifying tests.
+---
+
+# Test Writer Skill
+
+## Overview
+
+This skill provides guidelines for writing clean, maintainable Elixir/Phoenix tests following best practices. It focuses mainly on integration tests. For unit tests, we also want glanceable tests with unique names for values, test helpers, custom matchers and shared setups where appropriate.
+
+## Core Principles
+
+### 1. DRY Tests
+
+Avoid duplication by creating focused helper functions:
+
+**Bad:**
+```elixir
+test "GET /users returns index", %{conn: conn} do
+  conn = get(conn, "/users")
+  body = html_response(conn, 200)
+  assert body =~ "Users"
+end
+
+test "GET /users/:id returns show", %{conn: conn} do
+  conn = get(conn, "/users/1")
+  body = html_response(conn, 200)
+  assert body =~ "User"
+end
+```
+
+**Good:**
+```elixir
+defp goto_users_page(conn, suffix \\ ""), do: get(conn, "/users" <> suffix)
+
+test "GET /users returns index", %{conn: conn} do
+  conn = goto_users_page(conn)
+  assert html_response(conn, 200) =~ "Users"
+end
+
+test "GET /users/:id returns show", %{conn: conn} do
+  conn = goto_users_page(conn, "/1")
+  assert html_response(conn, 200) =~ "User"
+end
+```
+
+
+### 2. Separate Helpers for Different Assertion Patterns
+
+Don't use conditionals in helpers to handle different cases:
+
+**Bad:**
+```elixir
+defp goto_users_page(conn, suffix \\ "", check_title \\ true) do
+  path = "/users" <> suffix
+  conn = get(conn, path)
+  body = html_response(conn, 200)
+  if check_title, do: assert body =~ "Users"
+  assert body =~ "AppLayout"
+  body
+end
+```
+
+**Good:**
+```elixir
+defp goto_users_page(conn, suffix \\ "") do
+  path = "/users" <> suffix
+  conn = get(conn, path)
+  body = html_response(conn, 200)
+  assert body =~ "Users"
+  assert body =~ "AppLayout"
+  body
+end
+
+defp goto_user_page(conn, suffix) do
+  path = "/users" <> suffix
+  conn = get(conn, path)
+  body = html_response(conn, 200)
+  assert body =~ "AppLayout"
+  body
+end
+```
+
+## Test Structure
+
+### Context Block
+
+```elixir
+describe "resource name" do
+  # Shared setup in context if needed
+  # test "scenario" do ...
+end
+```
+
+### Value Aliasing
+
+Never reuse value names. Elixir is immutable, but value aliasing is confusing. Use unique, meaningful names for the left hand side of assignments. Or use pipes `|>` to eliminate the need for naming.
+
+```elixir
+test "GET /users returns index", %{conn: conn} do
+  # don't reassign
+  response = get(conn, "/users")
+  # ...
+end
+```
+
+## Common Patterns
+
+### HTML Pages with Layout
+
+```elixir
+defp goto_resource_page(conn, suffix \\ ""), do: ...
+# Asserts common layout elements and page-specific content
+```
+
+### API Endpoints
+
+```elixir
+test "GET /api/resource returns JSON", %{conn: conn} do
+  conn = conn |> put_req_header("accept", "application/json")
+  conn = get(conn, "/api/resource")
+  response = json_response(conn, 200)
+  # Assert structure
+end
+```
+
+### Error Handling
+
+```elixir
+test "returns 404 for nonexistent", %{conn: conn} do
+  assert html_response(get(conn, "/nonexistent"), 404)
+end
+```
+
+## Running Tests
+
+### One focused test file
+
+```bash
+cd /path/to/app
+mix test test/path/to_test.exs
+```
+
+### all tests
+```bash
+make test
+```

Dockerfile

@@ -0,0 +1,86 @@
+# Dockerfile for Dokku deployment
+# Multi-stage build for Phoenix/Elixir app with monorepo layout
+
+ARG ELIXIR_VERSION=1.18.3
+ARG OTP_VERSION=27.2.4
+ARG DEBIAN_VERSION=bookworm-20260316-slim
+
+ARG BUILDER_IMAGE="docker.io/hexpm/elixir:${ELIXIR_VERSION}-erlang-${OTP_VERSION}-debian-${DEBIAN_VERSION}"
+ARG RUNNER_IMAGE="docker.io/debian:${DEBIAN_VERSION}"
+
+# =============================================================================
+# Build stage
+# =============================================================================
+FROM ${BUILDER_IMAGE} AS builder
+
+RUN apt-get update -y && apt-get install -y build-essential git \
+    && apt-get clean && rm -f /var/lib/apt/lists/*_*
+
+WORKDIR /build
+
+# Install hex + rebar
+RUN mix local.hex --force && \
+    mix local.rebar --force
+
+ENV MIX_ENV="prod"
+
+# Copy blogex dependency first (changes less often)
+COPY blogex /build/blogex
+
+# Copy app dependency files first for better layer caching
+COPY app/mix.exs app/mix.lock /build/app/
+WORKDIR /build/app
+
+RUN mix deps.get --only $MIX_ENV
+RUN mkdir config
+
+# Copy compile-time config files
+COPY app/config/config.exs app/config/${MIX_ENV}.exs config/
+RUN mix deps.compile
+
+# Copy application source and compile
+COPY app/priv priv
+COPY app/assets assets
+COPY app/lib lib
+COPY app/rel rel
+COPY app/config/runtime.exs config/
+
+RUN mix compile
+
+# Build assets after compile (phoenix-colocated hooks need compiled app)
+RUN mix assets.deploy
+
+# Build the release
+RUN mix release
+
+# =============================================================================
+# Runtime stage
+# =============================================================================
+FROM ${RUNNER_IMAGE}
+
+RUN apt-get update -y && \
+    apt-get install -y libstdc++6 openssl libncurses5 locales ca-certificates \
+    && apt-get clean && rm -f /var/lib/apt/lists/*_*
+
+# Set the locale
+RUN sed -i '/en_US.UTF-8/s/^# //g' /etc/locale.gen && locale-gen
+ENV LANG=en_US.UTF-8
+ENV LANGUAGE=en_US:en
+ENV LC_ALL=en_US.UTF-8
+
+WORKDIR /app
+
+RUN chown nobody /app
+ENV MIX_ENV="prod"
+
+# Copy the release from the build stage
+COPY --from=builder --chown=nobody:root /build/app/_build/${MIX_ENV}/rel/firehose ./
+
+USER nobody
+
+# Dokku uses the EXPOSE port for routing
+EXPOSE 5000
+
+ENV PHX_SERVER=true
+
+CMD ["/app/bin/server"]

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Living Software LTD 
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

Makefile

@@ -0,0 +1,27 @@
+# Makefile for Firehose monorepo
+
+.PHONY: check precommit deps compile test format
+
+# Common check target that runs all static analysis
+check:
+	@echo "Running static analysis..."
+	@make -C app MISE_BIN=mise check
+
+# Precommit target for CI/pre-commit hooks
+precommit: check
+
+# Sync dependencies
+deps:
+	@make -C app deps
+
+# Compile the project
+compile:
+	@make -C app compile
+
+# Run tests
+test:
+	@make -C app test
+
+# Format code
+format:
+	@make -C app format

app.json

@@ -0,0 +1,20 @@
+{
+  "name": "firehose",
+  "healthchecks": {
+    "web": [
+      {
+        "type": "startup",
+        "name": "web check",
+        "path": "/",
+        "attempts": 5,
+        "wait": 3,
+        "timeout": 5
+      }
+    ]
+  },
+  "scripts": {
+    "dokku": {
+      "postdeploy": "/app/bin/migrate"
+    }
+  }
+}

app/.credo.exs

@@ -0,0 +1,226 @@
+# This file contains the configuration for Credo and you are probably reading
+# this after creating it with `mix credo.gen.config`.
+#
+# If you find anything wrong or unclear in this file, please report an
+# issue on GitHub: https://github.com/rrrene/credo/issues
+#
+%{
+  #
+  # You can have as many configs as you like in the `configs:` field.
+  configs: [
+    %{
+      #
+      # Run any config using `mix credo -C <name>`. If no config name is given
+      # "default" is used.
+      #
+      name: "default",
+      #
+      # These are the files included in the analysis:
+      files: %{
+        #
+        # You can give explicit globs or simply directories.
+        # In the latter case `**/*.{ex,exs}` will be used.
+        #
+        included: [
+          "lib/",
+          "src/",
+          "test/",
+          "web/",
+          "apps/*/lib/",
+          "apps/*/src/",
+          "apps/*/test/",
+          "apps/*/web/"
+        ],
+        excluded: [~r"/_build/", ~r"/deps/", ~r"/node_modules/"]
+      },
+      #
+      # Load and configure plugins here:
+      #
+      plugins: [],
+      #
+      # If you create your own checks, you must specify the source files for
+      # them here, so they can be loaded by Credo before running the analysis.
+      #
+      requires: ["lib_dev/firehose/checks/"],
+      #
+      # If you want to enforce a style guide and need a more traditional linting
+      # experience, you can change `strict` to `true` below:
+      #
+      strict: false,
+      #
+      # To modify the timeout for parsing files, change this value:
+      #
+      parse_timeout: 5000,
+      #
+      # If you want to use uncolored output by default, you can change `color`
+      # to `false` below:
+      #
+      color: true,
+      #
+      # You can customize the parameters of any check by adding a second element
+      # to the tuple.
+      #
+      # To disable a check put `false` as second element:
+      #
+      #     {Credo.Check.Design.DuplicatedCode, false}
+      #
+      checks: %{
+        enabled: [
+          #
+          ## Consistency Checks
+          #
+          {Credo.Check.Consistency.ExceptionNames, []},
+          {Credo.Check.Consistency.LineEndings, []},
+          {Credo.Check.Consistency.ParameterPatternMatching, []},
+          {Credo.Check.Consistency.SpaceAroundOperators, []},
+          {Credo.Check.Consistency.SpaceInParentheses, []},
+          {Credo.Check.Consistency.TabsOrSpaces, []},
+
+          #
+          ## Design Checks
+          #
+          # You can customize the priority of any check
+          # Priority values are: `low, normal, high, higher`
+          #
+          {Credo.Check.Design.AliasUsage,
+           [priority: :low, if_nested_deeper_than: 2, if_called_more_often_than: 0]},
+          {Credo.Check.Design.TagFIXME, []},
+          # You can also customize the exit_status of each check.
+          # If you don't want TODO comments to cause `mix credo` to fail, just
+          # set this value to 0 (zero).
+          #
+          {Credo.Check.Design.TagTODO, [exit_status: 2]},
+
+          #
+          ## Readability Checks
+          #
+          {Credo.Check.Readability.AliasOrder, []},
+          {Credo.Check.Readability.FunctionNames, []},
+          {Credo.Check.Readability.LargeNumbers, []},
+          {Credo.Check.Readability.MaxLineLength, [priority: :low, max_length: 120]},
+          {Credo.Check.Readability.ModuleAttributeNames, []},
+          {Credo.Check.Readability.ModuleDoc, []},
+          {Credo.Check.Readability.ModuleNames, []},
+          {Credo.Check.Readability.ParenthesesInCondition, []},
+          {Credo.Check.Readability.ParenthesesOnZeroArityDefs, []},
+          {Credo.Check.Readability.PipeIntoAnonymousFunctions, []},
+          {Credo.Check.Readability.PredicateFunctionNames, []},
+          {Credo.Check.Readability.PreferImplicitTry, []},
+          {Credo.Check.Readability.RedundantBlankLines, []},
+          {Credo.Check.Readability.Semicolons, []},
+          {Credo.Check.Readability.SpaceAfterCommas, []},
+          {Credo.Check.Readability.StringSigils, []},
+          {Credo.Check.Readability.TrailingBlankLine, []},
+          {Credo.Check.Readability.TrailingWhiteSpace, []},
+          {Credo.Check.Readability.UnnecessaryAliasExpansion, []},
+          {Credo.Check.Readability.VariableNames, []},
+          {Credo.Check.Readability.WithSingleClause, []},
+
+          #
+          ## Refactoring Opportunities
+          #
+          {Credo.Check.Refactor.Apply, []},
+          {Credo.Check.Refactor.CondStatements, []},
+          {Credo.Check.Refactor.CyclomaticComplexity, []},
+          {Credo.Check.Refactor.FilterCount, []},
+          {Credo.Check.Refactor.FilterFilter, []},
+          {Credo.Check.Refactor.FunctionArity, []},
+          {Credo.Check.Refactor.LongQuoteBlocks, []},
+          {Credo.Check.Refactor.MapJoin, []},
+          {Credo.Check.Refactor.MatchInCondition, []},
+          {Credo.Check.Refactor.NegatedConditionsInUnless, []},
+          {Credo.Check.Refactor.NegatedConditionsWithElse, []},
+          {Credo.Check.Refactor.Nesting, []},
+          {Credo.Check.Refactor.RedundantWithClauseResult, []},
+          {Credo.Check.Refactor.RejectReject, []},
+          {Credo.Check.Refactor.UnlessWithElse, []},
+          {Credo.Check.Refactor.WithClauses, []},
+
+          #
+          ## Warnings
+          #
+          {Credo.Check.Warning.ApplicationConfigInModuleAttribute, []},
+          {Credo.Check.Warning.BoolOperationOnSameValues, []},
+          {Credo.Check.Warning.Dbg, []},
+          {Credo.Check.Warning.ExpensiveEmptyEnumCheck, []},
+          {Credo.Check.Warning.IExPry, []},
+          {Credo.Check.Warning.IoInspect, []},
+          {Credo.Check.Warning.MissedMetadataKeyInLoggerConfig, []},
+          {Credo.Check.Warning.OperationOnSameValues, []},
+          {Credo.Check.Warning.OperationWithConstantResult, []},
+          {Credo.Check.Warning.RaiseInsideRescue, []},
+          {Credo.Check.Warning.SpecWithStruct, []},
+          {Credo.Check.Warning.StructFieldAmount, []},
+          {Credo.Check.Warning.UnsafeExec, []},
+          {Credo.Check.Warning.UnusedEnumOperation, []},
+          {Credo.Check.Warning.UnusedFileOperation, []},
+          {Credo.Check.Warning.UnusedKeywordOperation, []},
+          {Credo.Check.Warning.UnusedListOperation, []},
+          {Credo.Check.Warning.UnusedMapOperation, []},
+          {Credo.Check.Warning.UnusedPathOperation, []},
+          {Credo.Check.Warning.UnusedRegexOperation, []},
+          {Credo.Check.Warning.UnusedStringOperation, []},
+          {Credo.Check.Warning.UnusedTupleOperation, []},
+          {Credo.Check.Warning.WrongTestFilename, []},
+
+          #
+          ## Custom Checks
+          #
+          {Firehose.Checks.NoConnShadowing, []}
+        ],
+        disabled: [
+          #
+          # Checks scheduled for next check update (opt-in for now)
+          {Credo.Check.Refactor.UtcNowTruncate, []},
+
+          #
+          # Controversial and experimental checks (opt-in, just move the check to `:enabled`
+          #   and be sure to use `mix credo --strict` to see low priority checks)
+          #
+          {Credo.Check.Consistency.MultiAliasImportRequireUse, []},
+          {Credo.Check.Consistency.UnusedVariableNames, []},
+          {Credo.Check.Design.DuplicatedCode, []},
+          {Credo.Check.Design.SkipTestWithoutComment, []},
+          {Credo.Check.Readability.AliasAs, []},
+          {Credo.Check.Readability.BlockPipe, []},
+          {Credo.Check.Readability.ImplTrue, []},
+          {Credo.Check.Readability.MultiAlias, []},
+          {Credo.Check.Readability.NestedFunctionCalls, []},
+          {Credo.Check.Readability.OneArityFunctionInPipe, []},
+          {Credo.Check.Readability.OnePipePerLine, []},
+          {Credo.Check.Readability.SeparateAliasRequire, []},
+          {Credo.Check.Readability.SingleFunctionToBlockPipe, []},
+          {Credo.Check.Readability.SinglePipe, []},
+          {Credo.Check.Readability.Specs, []},
+          {Credo.Check.Readability.StrictModuleLayout, []},
+          {Credo.Check.Readability.WithCustomTaggedTuple, []},
+          {Credo.Check.Refactor.ABCSize, []},
+          {Credo.Check.Refactor.AppendSingleItem, []},
+          {Credo.Check.Refactor.CondInsteadOfIfElse, []},
+          {Credo.Check.Refactor.DoubleBooleanNegation, []},
+          {Credo.Check.Refactor.FilterReject, []},
+          {Credo.Check.Refactor.IoPuts, []},
+          {Credo.Check.Refactor.MapMap, []},
+          {Credo.Check.Refactor.ModuleDependencies, []},
+          {Credo.Check.Refactor.NegatedIsNil, []},
+          {Credo.Check.Refactor.PassAsyncInTestCases, []},
+          {Credo.Check.Refactor.PipeChainStart, []},
+          {Credo.Check.Refactor.RejectFilter, []},
+          {Credo.Check.Refactor.VariableRebinding, []},
+          {Credo.Check.Warning.LazyLogging, []},
+          {Credo.Check.Warning.LeakyEnvironment, []},
+          {Credo.Check.Warning.MapGetUnsafePass, []},
+          {Credo.Check.Warning.MixEnv, []},
+          {Credo.Check.Warning.UnsafeToAtom, []}
+          # {Credo.Check.Warning.UnusedOperation, [{MyMagicModule, [:fun1, :fun2]}]}
+
+          # {Credo.Check.Refactor.MapInto, []},
+
+          #
+          # Custom checks can be created using `mix credo.gen.check`.
+          #
+        ]
+      }
+    }
+  ]
+}

app/Makefile

@@ -0,0 +1,33 @@
+# Makefile for Firehose app
+
+MISE_BIN ?= /home/vscode/.local/bin/mise
+MISE_EXEC = $(MISE_BIN) exec --
+
+.PHONY: check precommit deps compile test format credo
+
+# Run all static analysis checks (no database required)
+check: credo format
+
+# Precommit target for CI/pre-commit hooks
+precommit: check compile
+
+# Sync dependencies
+deps:
+	$(MISE_EXEC) mix deps.get
+
+# Compile the project
+compile:
+	$(MISE_EXEC) mix compile --warnings-as-errors
+
+# Run tests (requires PostgreSQL running on localhost:5432)
+# Note: If you don't have PostgreSQL, you can skip tests with `make check`
+test: deps compile
+	$(MISE_EXEC) mix test
+
+# Format code
+format:
+	$(MISE_EXEC) mix format
+
+# Run Credo static analysis
+credo:
+	$(MISE_EXEC) mix credo --strict

app/config/config.exs

@@ -61,7 +61,8 @@ config :logger, :default_formatter,
 config :phoenix, :json_library, Jason
 
 config :blogex,
-  blogs: [Firehose.EngineeringBlog, Firehose.ReleaseNotes]
+  blogs: [Firehose.EngineeringBlog, Firehose.ReleaseNotes],
+  show_drafts: true
 
 # Import environment specific config. This must remain at the bottom
 # of this file so it overrides the configuration defined above.

app/config/dev.exs

@@ -4,7 +4,7 @@ import Config
 config :firehose, Firehose.Repo,
   username: "postgres",
   password: "postgres",
-  hostname: "localhost",
+  hostname: System.get_env("DB_HOST") || "localhost",
   database: "firehose_dev",
   stacktrace: true,
   show_sensitive_data_on_connection_error: true,

app/config/prod.exs

@@ -13,6 +13,9 @@ config :swoosh, api_client: Swoosh.ApiClient.Req
 # Disable Swoosh Local Memory Storage
 config :swoosh, local: false
 
+# Hide draft blog posts in production
+config :blogex, show_drafts: false
+
 # Do not print debug messages in production
 config :logger, level: :info
 

app/config/runtime.exs

@@ -51,7 +51,7 @@ if config_env() == :prod do
       """
 
   host = System.get_env("PHX_HOST") || "example.com"
-  port = String.to_integer(System.get_env("PORT") || "4000")
+  port = String.to_integer(System.get_env("PORT") || "5000")
 
   config :firehose, :dns_cluster_query, System.get_env("DNS_CLUSTER_QUERY")
 

app/config/test.exs

@@ -8,7 +8,7 @@ import Config
 config :firehose, Firehose.Repo,
   username: "postgres",
   password: "postgres",
-  hostname: "localhost",
+  hostname: System.get_env("DB_HOST") || "localhost",
   database: "firehose_test#{System.get_env("MIX_TEST_PARTITION")}",
   pool: Ecto.Adapters.SQL.Sandbox,
   pool_size: System.schedulers_online() * 2

app/lib/firehose/blogs/engineering_blog.ex

@@ -1,4 +1,7 @@
 defmodule Firehose.EngineeringBlog do
+  @moduledoc """
+  Engineering blog configuration.
+  """
   use Blogex.Blog,
     blog_id: :engineering,
     app: :firehose,

app/lib/firehose/blogs/release_notes.ex

@@ -1,4 +1,7 @@
 defmodule Firehose.ReleaseNotes do
+  @moduledoc """
+  Release notes blog configuration.
+  """
   use Blogex.Blog,
     blog_id: :release_notes,
     app: :firehose,

app/lib/firehose/release.ex

@@ -0,0 +1,31 @@
+defmodule Firehose.Release do
+  @moduledoc """
+  Tasks for production releases (e.g., database migrations).
+
+  Usage from Dokku:
+      dokku run APP_NAME /app/bin/migrate
+  """
+
+  @app :firehose
+
+  def migrate do
+    load_app()
+
+    for repo <- repos() do
+      {:ok, _, _} = Ecto.Migrator.with_repo(repo, &Ecto.Migrator.run(&1, :up, all: true))
+    end
+  end
+
+  def rollback(repo, version) do
+    load_app()
+    {:ok, _, _} = Ecto.Migrator.with_repo(repo, &Ecto.Migrator.run(&1, :down, to: version))
+  end
+
+  defp repos do
+    Application.fetch_env!(@app, :ecto_repos)
+  end
+
+  defp load_app do
+    Application.load(@app)
+  end
+end

app/lib/firehose_web.ex

@@ -88,8 +88,8 @@ defmodule FirehoseWeb do
       import FirehoseWeb.CoreComponents
 
       # Common modules used in templates
-      alias Phoenix.LiveView.JS
       alias FirehoseWeb.Layouts
+      alias Phoenix.LiveView.JS
 
       # Routes generation with the ~p sigil
       unquote(verified_routes())

app/lib/firehose_web/components/core_components.ex

@@ -29,6 +29,7 @@ defmodule FirehoseWeb.CoreComponents do
   use Phoenix.Component
   use Gettext, backend: FirehoseWeb.Gettext
 
+  alias Phoenix.HTML.Form
   alias Phoenix.LiveView.JS
 
   @doc """
@@ -181,7 +182,7 @@ defmodule FirehoseWeb.CoreComponents do
   def input(%{type: "checkbox"} = assigns) do
     assigns =
       assign_new(assigns, :checked, fn ->
-        Phoenix.HTML.Form.normalize_value("checkbox", assigns[:value])
+        Form.normalize_value("checkbox", assigns[:value])
       end)
 
     ~H"""

app/lib/firehose_web/components/layouts/app.html.heex

@@ -1,6 +1,9 @@
 <header class="navbar px-4 sm:px-6 lg:px-8 border-b border-base-200">
   <div class="flex-1">
-    <a href="/" class="font-display text-xl font-semibold tracking-tight text-primary hover:opacity-80 transition">
+    <a
+      href="/"
+      class="font-display text-xl font-semibold tracking-tight text-primary hover:opacity-80 transition"
+    >
       firehose
     </a>
   </div>

app/lib/firehose_web/controllers/blog_controller.ex

@@ -58,6 +58,7 @@ defmodule FirehoseWeb.BlogController do
   end
 
   defp parse_page(nil), do: 1
+
   defp parse_page(str) do
     case Integer.parse(str) do
       {page, ""} when page > 0 -> page

app/lib/firehose_web/controllers/blog_html/show.html.heex

@@ -1,4 +1,4 @@
 <div class="space-y-8">
   <a href={@base_path} class="text-sm text-primary hover:underline">&larr; Back to posts</a>
-  <.post_show post={@post} />
+  <.post_show post={@post} base_path={@base_path} />
 </div>

app/lib/firehose_web/controllers/blog_html/tag.html.heex

@@ -4,7 +4,7 @@
     <p class="mt-2 text-base-content/70">Posts tagged "{@tag}"</p>
   </header>
 
-  <.post_index posts={@posts} base_path={@base_path} />
+  <.post_index posts={@posts} base_path={@base_path} current_tag={@tag} />
 
   <a href={@base_path} class="text-sm text-primary hover:underline">&larr; All posts</a>
 </div>

app/lib/firehose_web/controllers/page_html/home.html.heex

@@ -6,7 +6,12 @@
     <div class="space-y-4 text-lg leading-relaxed text-base-content/80">
       <p>
         I'm <strong class="text-base-content">Willem van den Ende</strong>,
-        partner at <a href="https://qwan.eu" class="text-primary hover:underline" target="_blank" rel="noopener">QWAN</a>.
+        partner at <a
+          href="https://qwan.eu"
+          class="text-primary hover:underline"
+          target="_blank"
+          rel="noopener"
+        >QWAN</a>.
         This is where I write about AI-native consulting, shitty evals,
         and whatever prototype I'm building this week.
       </p>
@@ -21,7 +26,9 @@
         class="rounded-box border border-base-200 p-5 space-y-2 hover:border-primary/30 transition"
       >
         <a href={"#{post_base_path(post)}/#{post.id}"} class="block space-y-2">
-          <h3 class="font-semibold text-base-content hover:text-primary transition">{post.title}</h3>
+          <h3 class="font-semibold text-base-content hover:text-primary transition">
+            {post.title}
+          </h3>
           <p class="text-sm text-base-content/60">{post.description}</p>
           <div class="flex items-center gap-2 text-xs text-base-content/50">
             <time datetime={Date.to_iso8601(post.date)}>

app/lib_dev/firehose/checks/no_conn_shadowing.ex

@@ -0,0 +1,49 @@
+defmodule Firehose.Checks.NoConnShadowing do
+  use Credo.Check,
+    base_priority: :normal,
+    category: :readability,
+    explanations: [
+      check: """
+      Conn shadowing (`conn = get(conn, ...)`) makes Phoenix controller tests
+      noisy. Use pipe chains instead:
+
+          body = conn |> get("/path") |> html_response(200)
+
+      Run `./refactor_conn_aliasing.sh <file>` to fix automatically.
+      """
+    ]
+
+  @http_verbs ~w(get post put patch delete head options)a
+
+  @impl true
+  def run(%SourceFile{} = source_file, params) do
+    issue_meta = IssueMeta.for(source_file, params)
+
+    source_file
+    |> Credo.Code.prewalk(&traverse(&1, &2, issue_meta))
+    |> Enum.reverse()
+  end
+
+  defp traverse(
+         {:=, meta, [{:conn, _, _}, {verb, _, [{:conn, _, _} | _]}]} = ast,
+         issues,
+         issue_meta
+       )
+       when verb in @http_verbs do
+    issue = issue_for(issue_meta, meta[:line], verb)
+    {ast, [issue | issues]}
+  end
+
+  defp traverse(ast, issues, _issue_meta) do
+    {ast, issues}
+  end
+
+  defp issue_for(issue_meta, line_no, verb) do
+    format_issue(
+      issue_meta,
+      message:
+        "Conn shadowing detected (`conn = #{verb}(conn, ...)`). Run `./refactor_conn_aliasing.sh <file>` to fix.",
+      line_no: line_no
+    )
+  end
+end

app/mix.exs

@@ -32,7 +32,8 @@ defmodule Firehose.MixProject do
   end
 
   # Specifies which paths to compile per environment.
-  defp elixirc_paths(:test), do: ["lib", "test/support"]
+  defp elixirc_paths(:test), do: ["lib", "lib_dev", "test/support"]
+  defp elixirc_paths(:dev), do: ["lib", "lib_dev"]
   defp elixirc_paths(_), do: ["lib"]
 
   # Specifies your project dependencies.
@@ -66,7 +67,8 @@ defmodule Firehose.MixProject do
       {:jason, "~> 1.2"},
       {:dns_cluster, "~> 0.2.0"},
       {:bandit, "~> 1.5"},
-      {:blogex, path: "../blogex"}
+      {:blogex, path: "../blogex"},
+      {:credo, "~> 1.7", only: [:dev, :test], runtime: false}
     ]
   end
 

app/mix.lock

@@ -1,6 +1,8 @@
 %{
   "bandit": {:hex, :bandit, "1.10.3", "1e5d168fa79ec8de2860d1b4d878d97d4fbbe2fdbe7b0a7d9315a4359d1d4bb9", [:mix], [{:hpax, "~> 1.0", [hex: :hpax, repo: "hexpm", optional: false]}, {:plug, "~> 1.18", [hex: :plug, repo: "hexpm", optional: false]}, {:telemetry, "~> 0.4 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}, {:thousand_island, "~> 1.0", [hex: :thousand_island, repo: "hexpm", optional: false]}, {:websock, "~> 0.5", [hex: :websock, repo: "hexpm", optional: false]}], "hexpm", "99a52d909c48db65ca598e1962797659e3c0f1d06e825a50c3d75b74a5e2db18"},
+  "bunt": {:hex, :bunt, "1.0.0", "081c2c665f086849e6d57900292b3a161727ab40431219529f13c4ddcf3e7a44", [:mix], [], "hexpm", "dc5f86aa08a5f6fa6b8096f0735c4e76d54ae5c9fa2c143e5a1fc7c1cd9bb6b5"},
   "cc_precompiler": {:hex, :cc_precompiler, "0.1.11", "8c844d0b9fb98a3edea067f94f616b3f6b29b959b6b3bf25fee94ffe34364768", [:mix], [{:elixir_make, "~> 0.7", [hex: :elixir_make, repo: "hexpm", optional: false]}], "hexpm", "3427232caf0835f94680e5bcf082408a70b48ad68a5f5c0b02a3bea9f3a075b9"},
+  "credo": {:hex, :credo, "1.7.17", "f92b6aa5b26301eaa5a35e4d48ebf5aa1e7094ac00ae38f87086c562caf8a22f", [:mix], [{:bunt, "~> 0.2.1 or ~> 1.0", [hex: :bunt, repo: "hexpm", optional: false]}, {:file_system, "~> 0.2 or ~> 1.0", [hex: :file_system, repo: "hexpm", optional: false]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: false]}], "hexpm", "1eb5645c835f0b6c9b5410f94b5a185057bcf6d62a9c2b476da971cde8749645"},
   "db_connection": {:hex, :db_connection, "2.9.0", "a6a97c5c958a2d7091a58a9be40caf41ab496b0701d21e1d1abff3fa27a7f371", [:mix], [{:telemetry, "~> 0.4 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "17d502eacaf61829db98facf6f20808ed33da6ccf495354a41e64fe42f9c509c"},
   "decimal": {:hex, :decimal, "2.3.0", "3ad6255aa77b4a3c4f818171b12d237500e63525c2fd056699967a3e7ea20f62", [:mix], [], "hexpm", "a4d66355cb29cb47c3cf30e71329e58361cfcb37c34235ef3bf1d7bf3773aeac"},
   "dns_cluster": {:hex, :dns_cluster, "0.2.0", "aa8eb46e3bd0326bd67b84790c561733b25c5ba2fe3c7e36f28e88f384ebcb33", [:mix], [], "hexpm", "ba6f1893411c69c01b9e8e8f772062535a4cf70f3f35bcc964a324078d8c8240"},

app/priv/blog/engineering/2026/03-16-hello-world.md

@@ -1,6 +1,7 @@
 %{
   title: "Hello World",
   author: "Firehose Team",
+  published: false,
   tags: ~w(elixir phoenix),
   description: "Our first engineering blog post"
 }

app/priv/blog/engineering/2026/03-17-why-firehose.md

@@ -7,20 +7,24 @@
 }
 ---
 
-I wrote about [publishing short posts](https://www.qwan.eu/2025/05/20/publish-short-posts.html) on the QWAN blog last year. Giving myself license to write shorter, rougher pieces. That worked for a while. But some things don't belong on a consultancy blog.
+I wrote about [publishing short posts](https://www.qwan.eu/2025/05/20/publish-short-posts.html) on the QWAN blog last year. Giving myself license to write shorter, rougher pieces. That worked for a while. But some things don't feel like a good fit for the QWAN blog just yet.
 
-When I prototype with a coding agent at 11pm, the thing I learn is not a polished QWAN insight. It's a half-formed observation about evals, or a trick for keeping the human in the loop, or just "I built this and here's what surprised me." The QWAN blog has a certain standard. This stuff needs somewhere scruffier to land.
+This post was partially written with claude code, see the commit history [on our gitea](https://gitea.apps.sustainabledelivery.com/mostalive/firehose) if you want to check the differences.
 
-Hence **Firehose** — named after what it feels like to work with AI coding agents. You're drinking from a firehose of generated code, suggestions, and decisions. The interesting question is not "how do I generate more" but "how do I stay in control of what's coming out."
+When I prototype with a coding agent at 11pm (I should go to bed and write a post about sustainable pace the next day ;-) ), the thing I learn is not a polished QWAN insight. It's a half-formed observation about something that just happened, or a trick for keeping the human in the loop, or just "I built this and here's what surprised me." 
 
-That's also what this site is built with, by the way. The homepage, the blog engine, the layout — all built in conversation with Claude Code. I wanted to experience what our clients experience: shipping something real with an AI agent, and noticing where the friction is.
+Hence **Firehose** — named after what it feels like to work with AI coding agents. You're drinking from a firehose of generated code, suggestions, and decisions. The interesting question is not "how do I generate more" but "how do I stay in control of what's coming out.". And also, currently, how do I generate just enough and focus on interesting feedback loops instead of code? 
 
-A few things I noticed, building this:
+I wrote this last may as well [shallow research tool](https://www.qwan.eu/2025/05/01/agentic-search.html):
 
-- **Layout inheritance is a design decision.** The blog engine rendered pages outside Phoenix's layout pipeline. Getting navbar and CSS onto blog pages meant rethinking how the pieces fit together — not just adding a wrapper div.
-- **Warm aesthetics take intention.** The default Phoenix boilerplate is fine, but it says nothing about who you are. Choosing fonts and colours forced me to think about what "personal but professional" looks like.
-- **It's fast when it works, and confusing when it doesn't.** When the agent understands your stack, you move at extraordinary speed. When it doesn't (say, the difference between `@inner_block` and `@inner_content` in Phoenix layouts), you can burn time on a misunderstanding that a human would catch in seconds.
+> I want to both get better at using LLMs for programming, and also understand how they work. Marc suggested earlier this year that I write a series of blog posts about my use of them, but I have been drinking from a firehose, and it is quite difficult to figure out a good place to start writing.
+
+I have made good progress in learning, and at the same time, practices are still evolving. I see people write patterns. I think it is useful, but too early for that. I am at heuristics (rules of thumb).
+
+That is also why I open sourced the code for this blog [firehose repository on our gitea](https://gitea.apps.sustainabledelivery.com/mostalive/firehose). I think Jekyll, the static site generator we have for QWAN is passable, but I want the option to have a more interactive blog, and since this is going to be a firehose of ideas, give readers the option to subscribe to only what they are interested in, filter posts, like etc. I helped a friend with 'Ghost', but it felt clunky. I like writing in plain text and publishing with `git push` - that works with Jekyll and other static site generators.
+
+I am exploring working in small slices. That does require some initial investment in modularity. If you look at the code, you will notice that some of the blogging functionality is separate from the main site. I want an 'engineering blog' and 'release notes' as a plugin for Software as a Service applications. 
 
 This is the space I want to write in. Shorter than a conference talk, longer than a LinkedIn post. Honest about what works and what doesn't.
 
-If you're a CTO or engineering lead wondering what "AI-native development" actually looks like day to day — not the vendor pitch, the lived experience — that's what I'll be writing about here.
+If you're  wondering what "AI-native development" actually looks like day to day — not the vendor pitch, the lived experience — that's what I'll be writing about here.

app/priv/blog/engineering/2026/03-20-llm-simple-play.md

@@ -0,0 +1,12 @@
+%{
+  title: "Coding agent from scratch - a loop with tools, not that complicated",
+  author: "Willem van den Ende",
+  published: true,
+  tags: ~w(llm coding-agent python exercise),
+  description: "Coding agents are not that complicated. A loop with some tools. I found an interactive tutorial that lets you experience it"
+}
+---
+
+I had started on a "Write your own coding agent" exercise. Four iterations in, actually. And then I found [Tiny Agents]( https://tinyagents.dev/lesson/agent-loop), a set of interactive exercises that let you experience how agents work, from a simple chat request, through a tool, more tools etc. It has a live graph, that visualises of the flow of data and actions.
+
+It is good fun to play with, it starts simple and builds up. It lets you inspect the messages between the 'agent' loop code and the large language model server (which is just HTTP and some JSON).

app/priv/blog/engineering/2026/03-24-blog-triage.md

@@ -0,0 +1,46 @@
+%{
+  title: "Blog post triage with a local coding agent",
+  author: "Willem van den Ende",
+  published: true,
+  tags: ~w(llm coding-agent blogging),
+  description: "Can a coding agent help me get some of my draft blog posts over the line? I followed a tip by Chris Parsons to find out."
+}
+---
+
+I made a skill for a coding agent to help me get more of my draft blog posts over the line. I enjoy writing, and am somewhat fluent in it. Publishing that writing is more hit and miss, however. I often lose energy just before a piece is finished enough. I want to publish more often, and need to form a more effective habit for it.
+
+# What did I get out if it?
+
+I got a working agent 'skill' in an hour or so. I like the QWEN models for their no-bullshit approach to feedback. As it turns out, I have about 60 pages with the 'Candidate Blogpost' tag in my notes, but most of them are not more than an idea. Only some of them have enough detail to turn into a post. I am going to keep this around, prune my candidate blogposts, and add my recent clippings to the mix. 
+
+Quite a few of my candidates were 'just links' according to the model, but as I am inspired by [Simon Willison](https://www.simonwillison.net), there is value in sharing links with a brief description on why I think they are relevant. Probably in a different category.
+
+# How did I develop the skill? 
+
+I was inspired by two writings:
+
+- Jurgen De Smet asking [how do you write long form articles?](https://www.linkedin.com/posts/jurgendesmet_this-is-how-i-write-long-form-articles-these-share-7441394036222935040-JHmv). 
+- Chris Parsons suggested to [brief an agent for daily tasks](https://www.chrismdp.com/stop-prompting-start-briefing/), and use the _backbriefing_ loop from "The Art of Action" to improve them.
+
+I like "The Art of Action" - detailed, yet practical. So I had a chat with a frontier model to develop a skill for a local model to surface notes that are almost finished, with some suggestions to get them over the line.
+
+This was my initial prompt. Full chat transcript in the Further Reading section.
+
+#+begin_quote
+https://www.chrismdp.com/stop-prompting-start-briefing/ suggests an art of action style backbriefing loop for daily work. I would like to use a local model with pi, the shitty coding agent, instead of claude code. I have trouble publishing blogposts. I have many drafts, marked as CandidateBlogPost in an org-roam directory. I wonder if I could make some kind of pi extension or skill that finds candidate blogposts, helps identify ones that are almost finished, with a suggesion on what to do next for the top 3 almost finished, and suggestions for others on what to add. Probably prioritize recency. I could run that as a cron job in the morning, and create a new daily entry (I use daily entries for org-roam) to get me starte.d Goal would be not to have AI write my posts, but help me finish in pomodori instead of days.
+#+end_quote
+
+What I found interesting was that, maybe because I mentioned the links were in an sqlite database, claude desktop spontaneously suggested to create a bash script as part of the skill. I used to have a meta-skill to separate the deterministic parts of agent skills into scripts, but that does not seem to be necessary anymore. I prune my agent setups continuously, only keeping what is needed.
+
+# Tradeoffs
+
+Initially I planned to run this as a scheduled job, but from the development chat it emerged that backbriefing (improving the skill as we run it daily) would not work if it runs scheduled. 
+
+I chose a local coding agent with a local model, because I don't want to share my personal notes with a cloud service, and I thought that a smaller model would be more than powerful enough.
+
+
+## Further reading 
+
+https://claude.ai/share/be0184d9-f2bf-41ba-b2e3-235fe9daf9fd - initial chat do develop the skill
+
+I will share a repository with the skill later. I think it is more instructive to have a look at the prompt, and make one for your own notes, starting from your own goals.

app/rel/overlays/bin/migrate

@@ -0,0 +1,6 @@
+#!/bin/sh
+set -eu
+
+cd -P -- "$(dirname -- "$0")"/..
+
+exec ./bin/firehose eval Firehose.Release.migrate

app/rel/overlays/bin/server

@@ -0,0 +1,6 @@
+#!/bin/sh
+set -eu
+
+cd -P -- "$(dirname -- "$0")"/..
+
+PHX_SERVER=true exec ./bin/firehose start

app/test/firehose_web/controllers/blog_tags_test.exs

@@ -0,0 +1,123 @@
+defmodule FirehoseWeb.BlogTagsTest do
+  use FirehoseWeb.ConnCase
+
+  defp goto_engineering_tag_page(conn, tag) do
+    path = "/blog/engineering/tag/#{tag}"
+    conn_res = get(conn, path)
+    body = html_response(conn_res, 200)
+    assert body =~ ~s(tagged "#{tag}")
+    assert body =~ "Engineering Blog"
+    body
+  end
+
+  defp goto_releases_tag_page(conn, tag) do
+    path = "/blog/releases/tag/#{tag}"
+    conn_res = get(conn, path)
+    body = html_response(conn_res, 200)
+    assert body =~ ~s(tagged "#{tag}")
+    assert body =~ "Release Notes"
+    body
+  end
+
+  describe "engineering blog tags" do
+    test "GET /blog/engineering/tag/:tag shows tag page with all posts", %{conn: conn} do
+      body = goto_engineering_tag_page(conn, "elixir")
+      assert body =~ "Hello World"
+    end
+
+    test "GET /blog/engineering/tag/:tag page shows filtered posts", %{conn: conn} do
+      body = goto_engineering_tag_page(conn, "phoenix")
+      assert body =~ "Hello World"
+    end
+
+    test "GET /blog/engineering/tag/:tag page shows empty list for nonexistent tag", %{
+      conn: conn
+    } do
+      conn_res = get(conn, "/blog/engineering/tag/nonexistent-tag")
+      assert html_response(conn_res, 200) =~ ~s(tagged "nonexistent-tag")
+    end
+  end
+
+  describe "release notes blog tags" do
+    test "GET /blog/releases/tag/:tag shows tag page with all posts", %{conn: conn} do
+      body = goto_releases_tag_page(conn, "release")
+      assert body =~ "v0.1.0 Released"
+    end
+
+    test "GET /blog/releases/tag/:tag page shows filtered posts", %{conn: conn} do
+      conn_res = get(conn, "/blog/releases/tag/nonexistent-tag")
+      assert html_response(conn_res, 200) =~ ~s(tagged "nonexistent-tag")
+    end
+  end
+
+  describe "tag URL pattern" do
+    test "tag URLs follow pattern /blog/:blog_id/tag/:tag for engineering blog", %{conn: conn} do
+      # Test that the tag route exists and works correctly
+      conn_res1 = get(conn, "/blog/engineering/tag/elixir")
+      assert html_response(conn_res1, 200) =~ ~s(tagged "elixir")
+
+      conn_res2 = get(conn, "/blog/engineering/tag/phoenix")
+      assert html_response(conn_res2, 200) =~ ~s(tagged "phoenix")
+    end
+
+    test "tag URLs follow pattern /blog/:blog_id/tag/:tag for releases blog", %{conn: conn} do
+      # Test that the tag route exists and works correctly
+      conn_res = get(conn, "/blog/releases/tag/release")
+      assert html_response(conn_res, 200) =~ ~s(tagged "release")
+    end
+
+    test "nonexistent tags return 200 with empty post list", %{conn: conn} do
+      conn_res = get(conn, "/blog/engineering/tag/nonexistent-tag")
+      assert html_response(conn_res, 200)
+    end
+  end
+
+  describe "tag page structure" do
+    test "tag page has proper layout and back link", %{conn: conn} do
+      body = goto_engineering_tag_page(conn, "elixir")
+
+      assert body =~ "Engineering Blog"
+      assert body =~ ~s(tagged "elixir")
+      assert body =~ "All posts"
+    end
+
+    test "release tag page has proper layout and back link", %{conn: conn} do
+      body = goto_releases_tag_page(conn, "release")
+
+      assert body =~ "Release Notes"
+      assert body =~ ~s(tagged "release")
+      assert body =~ "All posts"
+    end
+  end
+
+  describe "clickable tags on index page" do
+    test "tags are rendered as clickable links on engineering blog index", %{
+      conn: conn
+    } do
+      conn_res1 = get(conn, "/blog/engineering")
+      body1 = html_response(conn_res1, 200)
+
+      # Verify tag links exist with correct href pattern
+      assert body1 =~ ~r{href="/blog/engineering/tag/meta"}
+      assert body1 =~ ~r{href="/blog/engineering/tag/ai"}
+    end
+
+    test "tags are rendered as clickable links on releases blog index", %{
+      conn: conn
+    } do
+      conn_res2 = get(conn, "/blog/releases")
+      body2 = html_response(conn_res2, 200)
+
+      # Verify tag link exists
+      assert body2 =~ ~r{href="/blog/releases/tag/release"}
+    end
+
+    test "tag links have proper styling classes", %{conn: conn} do
+      conn_res3 = get(conn, "/blog/engineering")
+      body3 = html_response(conn_res3, 200)
+
+      # Verify blogex-tag-link class is present for tag links
+      assert body3 =~ ~r{class="[^"]*blogex-tag-link}
+    end
+  end
+end

app/test/firehose_web/controllers/blog_test.exs

@@ -1,50 +1,54 @@
+# Firehose blog controller tests
+
 defmodule FirehoseWeb.BlogTest do
   use FirehoseWeb.ConnCase
 
+  defp visit_engineering_page(conn, suffix \\ "") do
+    path = "/blog/engineering" <> suffix
+    body = conn |> get(path) |> html_response(200)
+    assert body =~ "Engineering Blog"
+    assert body =~ "firehose"
+    body
+  end
+
+  defp visit_engineering_path(conn, suffix) do
+    path = "/blog/engineering" <> suffix
+    body = conn |> get(path) |> html_response(200)
+    assert body =~ "firehose"
+    body
+  end
+
   describe "engineering blog (HTML)" do
     test "GET /blog/engineering returns HTML index with layout", %{conn: conn} do
-      conn = get(conn, "/blog/engineering")
-      body = html_response(conn, 200)
-      assert body =~ "Engineering Blog"
-      assert body =~ "Hello World"
-      # Verify app layout is present (navbar)
-      assert body =~ "firehose"
+      visit_engineering_page(conn)
     end
 
     test "GET /blog/engineering/:slug returns HTML post with layout", %{conn: conn} do
-      conn = get(conn, "/blog/engineering/hello-world")
-      body = html_response(conn, 200)
+      body = visit_engineering_path(conn, "/hello-world")
       assert body =~ "Hello World"
-      assert body =~ "firehose"
     end
 
     test "GET /blog/engineering/tag/:tag returns HTML tag page", %{conn: conn} do
-      conn = get(conn, "/blog/engineering/tag/elixir")
-      body = html_response(conn, 200)
+      body = visit_engineering_path(conn, "/tag/elixir")
       assert body =~ ~s(tagged "elixir")
-      assert body =~ "Hello World"
     end
   end
 
   describe "input validation" do
     test "GET /blog/nonexistent returns 404", %{conn: conn} do
-      conn = get(conn, "/blog/nonexistent")
-      assert html_response(conn, 404)
+      assert conn |> get("/blog/nonexistent") |> html_response(404)
     end
 
     test "GET /blog/engineering?page=abc falls back to page 1", %{conn: conn} do
-      conn = get(conn, "/blog/engineering?page=abc")
-      assert html_response(conn, 200) =~ "Engineering Blog"
+      assert conn |> get("/blog/engineering?page=abc") |> html_response(200) =~ "Engineering Blog"
     end
 
     test "GET /blog/engineering?page=-1 falls back to page 1", %{conn: conn} do
-      conn = get(conn, "/blog/engineering?page=-1")
-      assert html_response(conn, 200) =~ "Engineering Blog"
+      assert conn |> get("/blog/engineering?page=-1") |> html_response(200) =~ "Engineering Blog"
     end
 
     test "GET /blog/engineering?page=0 falls back to page 1", %{conn: conn} do
-      conn = get(conn, "/blog/engineering?page=0")
-      assert html_response(conn, 200) =~ "Engineering Blog"
+      assert conn |> get("/blog/engineering?page=0") |> html_response(200) =~ "Engineering Blog"
     end
 
     test "GET /blog/engineering/nonexistent-post returns 404", %{conn: conn} do
@@ -56,61 +60,100 @@ defmodule FirehoseWeb.BlogTest do
 
   describe "release notes blog (HTML)" do
     test "GET /blog/releases returns HTML index", %{conn: conn} do
-      conn = get(conn, "/blog/releases")
-      body = html_response(conn, 200)
+      body = conn |> get("/blog/releases") |> html_response(200)
       assert body =~ "Release Notes"
       assert body =~ "v0.1.0 Released"
     end
 
     test "GET /blog/releases/:slug returns HTML post", %{conn: conn} do
-      conn = get(conn, "/blog/releases/v0-1-0")
-      body = html_response(conn, 200)
+      body = conn |> get("/blog/releases/v0-1-0") |> html_response(200)
       assert body =~ "v0.1.0 Released"
     end
+
+    test "GET /blog/releases/tag/:tag returns HTML tag page", %{conn: conn} do
+      body = conn |> get("/blog/releases/tag/elixir") |> html_response(200)
+      assert body =~ ~s(tagged "elixir")
+    end
   end
 
   describe "engineering blog (JSON API)" do
     test "GET /api/blog/engineering returns post index", %{conn: conn} do
-      conn = get(conn, "/api/blog/engineering")
-      assert %{"blog" => "engineering", "posts" => posts} = json_response(conn, 200)
+      assert %{"blog" => "engineering", "posts" => posts} =
+               conn
+               |> put_req_header("accept", "application/json")
+               |> get("/api/blog/engineering")
+               |> json_response(200)
+
       assert is_list(posts)
-      assert length(posts) > 0
+      refute Enum.empty?(posts)
     end
 
     test "GET /api/blog/engineering/:slug returns a post", %{conn: conn} do
-      conn = get(conn, "/api/blog/engineering/hello-world")
-      assert %{"id" => "hello-world", "title" => "Hello World"} = json_response(conn, 200)
+      assert %{"id" => "hello-world", "title" => "Hello World"} =
+               conn
+               |> put_req_header("accept", "application/json")
+               |> get("/api/blog/engineering/hello-world")
+               |> json_response(200)
     end
 
     test "GET /api/blog/engineering/:slug returns 404 for missing post", %{conn: conn} do
-      conn = get(conn, "/api/blog/engineering/nonexistent")
-      assert response(conn, 404)
+      assert conn
+             |> put_req_header("accept", "application/json")
+             |> get("/api/blog/engineering/nonexistent")
+             |> response(404)
     end
 
     test "GET /api/blog/engineering/feed.xml returns RSS", %{conn: conn} do
-      conn = get(conn, "/api/blog/engineering/feed.xml")
-      assert response_content_type(conn, :xml)
-      assert response(conn, 200) =~ "<rss"
+      response = conn |> get("/api/blog/engineering/feed.xml")
+      assert response(response, 200) =~ "<rss"
+      assert response_content_type(response, :xml)
+    end
+
+    test "GET /api/blog/engineering/tag/:tag returns JSON with posts", %{conn: conn} do
+      assert %{"blog" => "engineering", "tag" => "elixir", "posts" => posts} =
+               conn
+               |> put_req_header("accept", "application/json")
+               |> get("/api/blog/engineering/tag/elixir")
+               |> json_response(200)
+
+      assert is_list(posts)
     end
   end
 
   describe "release notes blog (JSON API)" do
     test "GET /api/blog/releases returns post index", %{conn: conn} do
-      conn = get(conn, "/api/blog/releases")
-      assert %{"blog" => "release_notes", "posts" => posts} = json_response(conn, 200)
+      assert %{"blog" => "release_notes", "posts" => posts} =
+               conn
+               |> put_req_header("accept", "application/json")
+               |> get("/api/blog/releases")
+               |> json_response(200)
+
       assert is_list(posts)
-      assert length(posts) > 0
+      refute Enum.empty?(posts)
     end
 
     test "GET /api/blog/releases/:slug returns a post", %{conn: conn} do
-      conn = get(conn, "/api/blog/releases/v0-1-0")
-      assert %{"id" => "v0-1-0", "title" => "v0.1.0 Released"} = json_response(conn, 200)
+      assert %{"id" => "v0-1-0", "title" => "v0.1.0 Released"} =
+               conn
+               |> put_req_header("accept", "application/json")
+               |> get("/api/blog/releases/v0-1-0")
+               |> json_response(200)
     end
 
     test "GET /api/blog/releases/feed.xml returns RSS", %{conn: conn} do
-      conn = get(conn, "/api/blog/releases/feed.xml")
-      assert response_content_type(conn, :xml)
-      assert response(conn, 200) =~ "<rss"
+      response = conn |> get("/api/blog/releases/feed.xml")
+      assert response(response, 200) =~ "<rss"
+      assert response_content_type(response, :xml)
+    end
+
+    test "GET /api/blog/releases/tag/:tag returns JSON with posts", %{conn: conn} do
+      assert %{"blog" => "release_notes", "tag" => "elixir", "posts" => posts} =
+               conn
+               |> put_req_header("accept", "application/json")
+               |> get("/api/blog/releases/tag/elixir")
+               |> json_response(200)
+
+      assert is_list(posts)
     end
   end
 end

app/test/firehose_web/controllers/page_controller_test.exs

@@ -2,8 +2,7 @@ defmodule FirehoseWeb.PageControllerTest do
   use FirehoseWeb.ConnCase
 
   test "GET /", %{conn: conn} do
-    conn = get(conn, ~p"/")
-    body = html_response(conn, 200)
+    body = conn |> get(~p"/") |> html_response(200)
     assert body =~ "Drinking from the firehose"
     assert body =~ "Willem van den Ende"
   end

app/test/support/data_case.ex

@@ -14,6 +14,8 @@ defmodule Firehose.DataCase do
   this option is not recommended for other databases.
   """
 
+  alias Ecto.Adapters.SQL.Sandbox
+
   use ExUnit.CaseTemplate
 
   using do
@@ -36,8 +38,8 @@ defmodule Firehose.DataCase do
   Sets up the sandbox based on the test tags.
   """
   def setup_sandbox(tags) do
-    pid = Ecto.Adapters.SQL.Sandbox.start_owner!(Firehose.Repo, shared: not tags[:async])
-    on_exit(fn -> Ecto.Adapters.SQL.Sandbox.stop_owner(pid) end)
+    pid = Sandbox.start_owner!(Firehose.Repo, shared: not tags[:async])
+    on_exit(fn -> Sandbox.stop_owner(pid) end)
   end
 
   @doc """

blogex/lib/blogex.ex

@@ -110,6 +110,11 @@ defmodule Blogex do
     * `Blogex.Router` — mountable Plug router
   """
 
+  @doc "Returns true if draft posts should be visible (dev/test environments)."
+  def show_drafts? do
+    Application.get_env(:blogex, :show_drafts, false)
+  end
+
   defdelegate blogs, to: Blogex.Registry
   defdelegate get_blog!(blog_id), to: Blogex.Registry
   defdelegate get_blog(blog_id), to: Blogex.Registry

blogex/lib/blogex/blog.ex

@@ -73,8 +73,14 @@ defmodule Blogex.Blog do
       @doc "Returns the base URL path for this blog."
       def base_path, do: @blog_base_path
 
-      @doc "Returns all published posts, newest first."
-      def all_posts, do: Enum.filter(@posts, & &1.published)
+      @doc "Returns all visible posts, newest first. Drafts are included in dev/test."
+      def all_posts do
+        if Blogex.show_drafts?() do
+          @posts
+        else
+          Enum.filter(@posts, & &1.published)
+        end
+      end
 
       @doc "Returns the N most recent published posts."
       def recent_posts(n \\ 5), do: Enum.take(all_posts(), n)

blogex/lib/blogex/components.ex

@@ -23,9 +23,11 @@ defmodule Blogex.Components do
 
     * `:posts` - list of `%Blogex.Post{}` structs (required)
     * `:base_path` - base URL path for post links (required)
+    * `:current_tag` - currently selected tag for highlighting (optional)
   """
   attr :posts, :list, required: true
   attr :base_path, :string, required: true
+  attr :current_tag, :string, default: nil
 
   def post_index(assigns) do
     ~H"""
@@ -35,7 +37,7 @@ defmodule Blogex.Components do
           <h2>
             <a href={"#{@base_path}/#{post.id}"}>{post.title}</a>
           </h2>
-          <.post_meta post={post} />
+          <.post_meta post={post} base_path={@base_path} current_tag={@current_tag} />
         </header>
         <p class="blogex-post-description">{post.description}</p>
       </article>
@@ -49,15 +51,17 @@ defmodule Blogex.Components do
   ## Attributes
 
     * `:post` - a `%Blogex.Post{}` struct (required)
+    * `:base_path` - base URL path for tag links (required)
   """
   attr :post, :map, required: true
+  attr :base_path, :string, required: true
 
   def post_show(assigns) do
     ~H"""
     <article class="blogex-post">
       <header class="blogex-post-header">
         <h1>{@post.title}</h1>
-        <.post_meta post={@post} />
+        <.post_meta post={@post} base_path={@base_path} />
       </header>
       <div class="blogex-post-body">
         {Phoenix.HTML.raw(@post.body)}
@@ -68,8 +72,16 @@ defmodule Blogex.Components do
 
   @doc """
   Renders post metadata (date, author, tags).
+
+  ## Attributes
+
+    * `:post` - a `%Blogex.Post{}` struct (required)
+    * `:base_path` - base URL path for tag links (required)
+    * `:current_tag` - currently selected tag for highlighting (optional)
   """
   attr :post, :map, required: true
+  attr :base_path, :string, required: true
+  attr :current_tag, :string, default: nil
 
   def post_meta(assigns) do
     ~H"""
@@ -80,9 +92,13 @@ defmodule Blogex.Components do
       <span :if={@post.author} class="blogex-post-author">
         by {@post.author}
       </span>
-      <span :for={tag <- @post.tags} class="blogex-tag">
+      <a
+        :for={tag <- @post.tags}
+        href={"#{@base_path}/tag/#{tag}"}
+        class={["blogex-tag-link", tag == @current_tag && "blogex-tag-active"]}
+      >
         {tag}
-      </span>
+      </a>
     </div>
     """
   end

blogex/lib/blogex/layout.ex

@@ -64,7 +64,7 @@ defmodule Blogex.Layout do
       </head>
       <body style="max-width: 48rem; margin: 0 auto; padding: 2rem; font-family: system-ui, sans-serif;">
         <nav><a href={@base_path}>&larr; Back</a></nav>
-        <.post_show post={@post} />
+        <.post_show post={@post} base_path={@base_path} />
       </body>
     </html>
     """

blogex/mix.lock

@@ -0,0 +1,22 @@
+%{
+  "earmark": {:hex, :earmark, "1.4.48", "5f41e579d85ef812351211842b6e005f6e0cef111216dea7d4b9d58af4608434", [:mix], [], "hexpm", "a461a0ddfdc5432381c876af1c86c411fd78a25790c75023c7a4c035fdc858f9"},
+  "earmark_parser": {:hex, :earmark_parser, "1.4.44", "f20830dd6b5c77afe2b063777ddbbff09f9759396500cdbe7523efd58d7a339c", [:mix], [], "hexpm", "4778ac752b4701a5599215f7030989c989ffdc4f6df457c5f36938cc2d2a2750"},
+  "ex_doc": {:hex, :ex_doc, "0.40.1", "67542e4b6dde74811cfd580e2c0149b78010fd13001fda7cfeb2b2c2ffb1344d", [:mix], [{:earmark_parser, "~> 1.4.44", [hex: :earmark_parser, repo: "hexpm", optional: false]}, {:makeup_c, ">= 0.1.0", [hex: :makeup_c, repo: "hexpm", optional: true]}, {:makeup_elixir, "~> 0.14 or ~> 1.0", [hex: :makeup_elixir, repo: "hexpm", optional: false]}, {:makeup_erlang, "~> 0.1 or ~> 1.0", [hex: :makeup_erlang, repo: "hexpm", optional: false]}, {:makeup_html, ">= 0.1.0", [hex: :makeup_html, repo: "hexpm", optional: true]}], "hexpm", "bcef0e2d360d93ac19f01a85d58f91752d930c0a30e2681145feea6bd3516e00"},
+  "jason": {:hex, :jason, "1.4.4", "b9226785a9aa77b6857ca22832cffa5d5011a667207eb2a0ad56adb5db443b8a", [:mix], [{:decimal, "~> 1.0 or ~> 2.0", [hex: :decimal, repo: "hexpm", optional: true]}], "hexpm", "c5eb0cab91f094599f94d55bc63409236a8ec69a21a67814529e8d5f6cc90b3b"},
+  "makeup": {:hex, :makeup, "1.2.1", "e90ac1c65589ef354378def3ba19d401e739ee7ee06fb47f94c687016e3713d1", [:mix], [{:nimble_parsec, "~> 1.4", [hex: :nimble_parsec, repo: "hexpm", optional: false]}], "hexpm", "d36484867b0bae0fea568d10131197a4c2e47056a6fbe84922bf6ba71c8d17ce"},
+  "makeup_elixir": {:hex, :makeup_elixir, "1.0.1", "e928a4f984e795e41e3abd27bfc09f51db16ab8ba1aebdba2b3a575437efafc2", [:mix], [{:makeup, "~> 1.0", [hex: :makeup, repo: "hexpm", optional: false]}, {:nimble_parsec, "~> 1.2.3 or ~> 1.3", [hex: :nimble_parsec, repo: "hexpm", optional: false]}], "hexpm", "7284900d412a3e5cfd97fdaed4f5ed389b8f2b4cb49efc0eb3bd10e2febf9507"},
+  "makeup_erlang": {:hex, :makeup_erlang, "1.0.3", "4252d5d4098da7415c390e847c814bad3764c94a814a0b4245176215615e1035", [:mix], [{:makeup, "~> 1.0", [hex: :makeup, repo: "hexpm", optional: false]}], "hexpm", "953297c02582a33411ac6208f2c6e55f0e870df7f80da724ed613f10e6706afd"},
+  "mime": {:hex, :mime, "2.0.7", "b8d739037be7cd402aee1ba0306edfdef982687ee7e9859bee6198c1e7e2f128", [:mix], [], "hexpm", "6171188e399ee16023ffc5b76ce445eb6d9672e2e241d2df6050f3c771e80ccd"},
+  "nimble_parsec": {:hex, :nimble_parsec, "1.4.2", "8efba0122db06df95bfaa78f791344a89352ba04baedd3849593bfce4d0dc1c6", [:mix], [], "hexpm", "4b21398942dda052b403bbe1da991ccd03a053668d147d53fb8c4e0efe09c973"},
+  "nimble_publisher": {:hex, :nimble_publisher, "1.1.1", "3ea4d4cfca45b11a5377bce7608367a9ddd7e717a9098161d8439eca23e239aa", [:mix], [{:earmark, "~> 1.4", [hex: :earmark, repo: "hexpm", optional: false]}, {:makeup, "~> 1.0", [hex: :makeup, repo: "hexpm", optional: false]}], "hexpm", "d67e15bddf07e8c60f75849008b78ea8c6b2b4ae8e3f882ccf0a22d57bd42ed0"},
+  "phoenix": {:hex, :phoenix, "1.8.5", "919db335247e6d4891764dc3063415b0d2457641c5f9b3751b5df03d8e20bbcf", [:mix], [{:bandit, "~> 1.0", [hex: :bandit, repo: "hexpm", optional: true]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: true]}, {:phoenix_pubsub, "~> 2.1", [hex: :phoenix_pubsub, repo: "hexpm", optional: false]}, {:phoenix_template, "~> 1.0", [hex: :phoenix_template, repo: "hexpm", optional: false]}, {:phoenix_view, "~> 2.0", [hex: :phoenix_view, repo: "hexpm", optional: true]}, {:plug, "~> 1.14", [hex: :plug, repo: "hexpm", optional: false]}, {:plug_cowboy, "~> 2.7", [hex: :plug_cowboy, repo: "hexpm", optional: true]}, {:plug_crypto, "~> 1.2 or ~> 2.0", [hex: :plug_crypto, repo: "hexpm", optional: false]}, {:telemetry, "~> 0.4 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}, {:websock_adapter, "~> 0.5.3", [hex: :websock_adapter, repo: "hexpm", optional: false]}], "hexpm", "83b2bb125127e02e9f475c8e3e92736325b5b01b0b9b05407bcb4083b7a32485"},
+  "phoenix_html": {:hex, :phoenix_html, "4.3.0", "d3577a5df4b6954cd7890c84d955c470b5310bb49647f0a114a6eeecc850f7ad", [:mix], [], "hexpm", "3eaa290a78bab0f075f791a46a981bbe769d94bc776869f4f3063a14f30497ad"},
+  "phoenix_live_view": {:hex, :phoenix_live_view, "1.1.27", "9afcab28b0c82afdc51044e661bcd5b8de53d242593d34c964a37710b40a42af", [:mix], [{:igniter, ">= 0.6.16 and < 1.0.0-0", [hex: :igniter, repo: "hexpm", optional: true]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: true]}, {:lazy_html, "~> 0.1.0", [hex: :lazy_html, repo: "hexpm", optional: true]}, {:phoenix, "~> 1.6.15 or ~> 1.7.0 or ~> 1.8.0-rc", [hex: :phoenix, repo: "hexpm", optional: false]}, {:phoenix_html, "~> 3.3 or ~> 4.0", [hex: :phoenix_html, repo: "hexpm", optional: false]}, {:phoenix_template, "~> 1.0", [hex: :phoenix_template, repo: "hexpm", optional: false]}, {:phoenix_view, "~> 2.0", [hex: :phoenix_view, repo: "hexpm", optional: true]}, {:plug, "~> 1.15", [hex: :plug, repo: "hexpm", optional: false]}, {:telemetry, "~> 0.4.2 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "415735d0b2c612c9104108b35654e977626a0cb346711e1e4f1ed16e3c827ede"},
+  "phoenix_pubsub": {:hex, :phoenix_pubsub, "2.2.0", "ff3a5616e1bed6804de7773b92cbccfc0b0f473faf1f63d7daf1206c7aeaaa6f", [:mix], [], "hexpm", "adc313a5bf7136039f63cfd9668fde73bba0765e0614cba80c06ac9460ff3e96"},
+  "phoenix_template": {:hex, :phoenix_template, "1.0.4", "e2092c132f3b5e5b2d49c96695342eb36d0ed514c5b252a77048d5969330d639", [:mix], [{:phoenix_html, "~> 2.14.2 or ~> 3.0 or ~> 4.0", [hex: :phoenix_html, repo: "hexpm", optional: true]}], "hexpm", "2c0c81f0e5c6753faf5cca2f229c9709919aba34fab866d3bc05060c9c444206"},
+  "plug": {:hex, :plug, "1.19.1", "09bac17ae7a001a68ae393658aa23c7e38782be5c5c00c80be82901262c394c0", [:mix], [{:mime, "~> 1.0 or ~> 2.0", [hex: :mime, repo: "hexpm", optional: false]}, {:plug_crypto, "~> 1.1.1 or ~> 1.2 or ~> 2.0", [hex: :plug_crypto, repo: "hexpm", optional: false]}, {:telemetry, "~> 0.4.3 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "560a0017a8f6d5d30146916862aaf9300b7280063651dd7e532b8be168511e62"},
+  "plug_crypto": {:hex, :plug_crypto, "2.1.1", "19bda8184399cb24afa10be734f84a16ea0a2bc65054e23a62bb10f06bc89491", [:mix], [], "hexpm", "6470bce6ffe41c8bd497612ffde1a7e4af67f36a15eea5f921af71cf3e11247c"},
+  "telemetry": {:hex, :telemetry, "1.4.1", "ab6de178e2b29b58e8256b92b382ea3f590a47152ca3651ea857a6cae05ac423", [:rebar3], [], "hexpm", "2172e05a27531d3d31dd9782841065c50dd5c3c7699d95266b2edd54c2dafa1c"},
+  "websock": {:hex, :websock, "0.5.3", "2f69a6ebe810328555b6fe5c831a851f485e303a7c8ce6c5f675abeb20ebdadc", [:mix], [], "hexpm", "6105453d7fac22c712ad66fab1d45abdf049868f253cf719b625151460b8b453"},
+  "websock_adapter": {:hex, :websock_adapter, "0.5.9", "43dc3ba6d89ef5dec5b1d0a39698436a1e856d000d84bf31a3149862b01a287f", [:mix], [{:bandit, ">= 0.6.0", [hex: :bandit, repo: "hexpm", optional: true]}, {:plug, "~> 1.14", [hex: :plug, repo: "hexpm", optional: false]}, {:plug_cowboy, "~> 2.6", [hex: :plug_cowboy, repo: "hexpm", optional: true]}, {:websock, "~> 0.5", [hex: :websock, repo: "hexpm", optional: false]}], "hexpm", "5534d5c9adad3c18a0f58a9371220d75a803bf0b9a3d87e6fe072faaeed76a08"},
+}

blogex/priv/blog/engineering/2026/02-15-testing-liveview-at-scale.md

@@ -5,6 +5,8 @@
   description: "Our testing strategy for 200+ LiveView modules"
 }
 ---
+*This is a sample blog post, generated to show what blogex can do.*
+
 With over 200 LiveView modules in our codebase, we needed a testing strategy
 that was both fast and reliable. Here's what we landed on.
 

blogex/priv/blog/engineering/2026/03-10-rebuilding-data-pipeline.md

@@ -5,6 +5,8 @@
   description: "How we replaced our Kafka consumer with Broadway for 10x throughput"
 }
 ---
+*This is a sample blog post, generated to show what blogex can do.*
+
 Last quarter we hit a wall with our homegrown Kafka consumer. Message lag was
 growing, backpressure was non-existent, and our on-call engineers were losing
 sleep. We decided to rebuild on [Broadway](https://github.com/dashbitco/broadway).

blogex/priv/blog/release-notes/2026/02-01-v2-3-0.md

@@ -5,6 +5,8 @@
   description: "Reliable webhook delivery, dark mode, and improved search"
 }
 ---
+*This is a sample blog post, generated to show what blogex can do.*
+
 Here's what landed in v2.3.0.
 
 ## Webhook Reliability

blogex/priv/blog/release-notes/2026/03-01-v2-4-0.md

@@ -5,6 +5,8 @@
   description: "New team dashboards, API rate limiting, and 12 bug fixes"
 }
 ---
+*This is a sample blog post, generated to show what blogex can do.*
+
 We're excited to ship v2.4.0 with two major features and a pile of bug fixes.
 
 ## Team Dashboards

context.md

@@ -1,172 +1,7 @@
-# Code Context
+Investigation complete. Found the tag implementation details:
 
-## Files Retrieved
-List with exact line ranges:
-1. `app/test/firehose_web/controllers/blog_test.exs` (lines 1-128) - Comprehensive blog controller tests covering HTML and JSON API endpoints for engineering blog and release notes
-2. `app/lib/firehose_web/controllers/blog_controller.ex` (lines 1-79) - Blog controller with pagination, 404 handling, and input validation
-3. `app/test/support/conn_case.ex` (lines 1-38) - Test case template for connection tests
-4. `app/lib/firehose/blogs/engineering_blog.ex` (lines 1-7) - Engineering blog module configuration
-5. `app/lib/firehose/blogs/release_notes.ex` (lines 1-7) - Release notes blog module configuration
+**Key Finding**: The `post_meta` component in `/workspaces/firehose/blogex/lib/blogex/components.ex` (lines 83-85) renders tags as plain text without links, while there's already a working `tag_list` component (lines 93-115) that properly creates links with the pattern `href={"#{@base_path}/tag/#{tag}"}`.
 
-## Key Code
+**Route structure**: `/tag/:tag` in `/workspaces/firehose/blogex/lib/blogex/router.ex` (line 62) handles tag filtering via `blog.posts_by_tag(tag)`.
 
-### Test Organization
-```elixir
-# Current structure has 4 describe blocks:
-describe "engineering blog (HTML)"     # 3 tests
-describe "input validation"           # 5 tests (newly added in last commit)
-describe "release notes blog (HTML)"  # 2 tests
-describe "engineering blog (JSON API)" # 4 tests  
-describe "release notes blog (JSON API)" # 3 tests
-```
-
-### Input Validation Logic (blog_controller.ex, lines 68-76)
-```elixir
-defp parse_page(nil), do: 1
-defp parse_page(str) do
-  case Integer.parse(str) do
-    {page, ""} when page > 0 -> page
-    _ -> 1
-  end
-end
-```
-
-### Test Coverage Added in Last Commit
-```elixir
-describe "input validation" do
-  test "GET /blog/nonexistent returns 404", %{conn: conn} do
-    conn = get(conn, "/blog/nonexistent")
-    assert html_response(conn, 404)
-  end
-
-  test "GET /blog/engineering?page=abc falls back to page 1", %{conn: conn} do
-    conn = get(conn, "/blog/engineering?page=abc")
-    assert html_response(conn, 200) =~ "Engineering Blog"
-  end
-
-  test "GET /blog/engineering?page=-1 falls back to page 1", %{conn: conn} do
-    conn = get(conn, "/blog/engineering?page=-1")
-    assert html_response(conn, 200) =~ "Engineering Blog"
-  end
-
-  test "GET /blog/engineering?page=0 falls back to page 1", %{conn: conn} do
-    conn = get(conn, "/blog/engineering?page=0")
-    assert html_response(conn, 200) =~ "Engineering Blog"
-  end
-
-  test "GET /blog/engineering/nonexistent-post returns 404", %{conn: conn} do
-    assert_raise Blogex.NotFoundError, fn ->
-      get(conn, "/blog/engineering/nonexistent-post")
-    end
-  end
-end
-```
-
-## Architecture
-The application uses:
-- **Blogex** library for blog functionality (engineering blog and release notes)
-- **Phoenix** framework for web endpoints
-- **ConnCase** test helper for connection testing
-- Two blog types: `Firehose.EngineeringBlog` and `Firehose.ReleaseNotes`
-- Pagination through `blog.paginate(page)` method
-- 404 handling via `Blogex.NotFoundError` exception
-
-## Start Here
-Which file to look at first and why:
-
-**Start with `app/lib/firehose_web/controllers/blog_controller.ex`**
-
-Why: This is the central controller that handles all blog requests. Understanding its structure (especially the `parse_page/1` function and `resolve_blog/2` plug) provides context for why the validation tests were added and how input handling works across both HTML and JSON endpoints.
-
-## Code Smells & Refactoring Suggestions
-
-### Smell 1: Repetitive Validation Tests
-**Issue**: Four tests for page parameter validation (`page=abc`, `-1`, `0`, and valid values) are highly repetitive with identical assertions.
-
-**Refactoring Suggestion**: Use parameterized tests or test helpers:
-```elixir
-# Test helper approach
-test_page_fallback("page=abc", "abc")
-test_page_fallback("page=-1", "-1") 
-test_page_fallback("page=0", "0")
-
-defp test_page_fallback(query_param, expected_page) do
-  conn = get(conn, "/blog/engineering?#{query_param}")
-  assert html_response(conn, 200) =~ "Engineering Blog"
-end
-```
-
-### Smell 2: Missing Negative Test Coverage
-**Issue**: Tests don't verify what happens when invalid blog_id is provided (e.g., `/blog/invalid-blog`).
-
-**Refactoring Suggestion**: Add test for unknown blog:
-```elixir
-test "GET /blog/unknown returns 404", %{conn: conn} do
-  conn = get(conn, "/blog/unknown")
-  assert html_response(conn, 404)
-end
-```
-
-### Smell 3: Inconsistent Test Naming
-**Issue**: Some tests use hyphenated slugs (`v0-1-0`), others use different formats. The naming doesn't clearly indicate what's being tested.
-
-**Refactoring Suggestion**: Standardize naming:
-```elixir
-# Instead of: "GET /blog/releases/v0-1-0 returns HTML post"
-test "GET /blog/releases/:slug returns a release post", %{conn: conn} do
-```
-
-### Smell 4: Redundant Layout Assertions
-**Issue**: Multiple tests assert the same "firehose" string appears in response, testing layout presence.
-
-**Refactoring Suggestion**: Create a shared test helper:
-```elixir
-defp assert_has_app_layout(body), 
-  do: assert body =~ "firehose"
-
-# Then in tests: assert_has_app_layout(body)
-```
-
-### Smell 5: Test Order Doesn't Follow Flow
-**Issue**: Tests are grouped by endpoint but validation tests (which should be first for defensive programming) are in the middle.
-
-**Refactoring Suggestion**: Reorder to follow natural request flow:
-1. Input validation (404s, invalid params)
-2. Success cases (index, show, tag)
-3. Edge cases (pagination, RSS feeds)
-
-### Smell 6: No Test for Controller-Level Error Handling
-**Issue**: The controller uses `halt()` in the resolve_blog plug, but there's no test verifying this behavior.
-
-**Refactoring Suggestion**: Add test:
-```elixir
-test "GET /blog/:blog_id with invalid blog halts request", %{conn: conn} do
-  conn = get(conn, "/blog/invalid")
-  assert conn.halted
-end
-```
-
-### Smell 7: Mixed Response Types Without Clear Separation
-**Issue**: HTML tests use `html_response/2`, JSON tests use `json_response/2`, but there's no helper to verify content type before parsing.
-
-**Refactoring Suggestion**: Create response helpers:
-```elixir
-defp assert_html(conn, status), do: assert html_response(conn, status) != ""
-defp assert_json(conn, status), do: assert json_response(conn, status) != %{}
-```
-
-### Smell 8: No Test for Concurrent Requests or Edge Cases
-**Issue**: Missing tests for:
-- Empty page parameter (`?page=`)
-- Very large page numbers
-- Special characters in slug/tag parameters
-
-**Refactoring Suggestion**: Add edge case tests to validation describe block.
-
-### Overall Recommendations
-1. **Extract test helpers** to reduce duplication (especially for page validation)
-2. **Standardize test naming** conventions across all blog types
-3. **Add positive test** for valid page numbers (currently missing)
-4. **Consider property-based testing** for input validation scenarios
-5. **Add performance tests** if pagination is used heavily
-6. **Create integration tests** that verify end-to-end flows
+**Tests exist**: `/workspaces/firehose/app/test/firehose_web/controllers/blog_test.exs` (lines 35-42, 117-122) verify tag page functionality.

new-post.md

@@ -0,0 +1,45 @@
+```mermaid
+sequenceDiagram
+    participant User
+    participant Engineering as Engineering Folder<br/>(priv/blog/engineering)
+    participant Blogex as Blogex Library
+    participant PhoenixApp as Firehose Web App
+    participant Browser
+
+    Note over User,Browser: New Markdown File Flow
+
+    User->>Engineering: Create markdown file<br/>(e.g., new-post.md)
+    
+    Note over Engineering: File appears in directory
+    
+    Note over Blogex: Blogex reads markdown files at app startup<br/>via config (priv/blog/engineering/**/*.md)
+    
+    PhoenixApp->>Blogex: Request post index via BlogController<br/>(GET /blog/engineering)
+    Blogex->>Engineering: Read markdown files from priv/blog/engineering/
+    Blogex->>Blogex: Parse markdown + frontmatter
+    Blogex->>Blogex: Create %Blogex.Post{ structs}
+    
+    Note over Blogex: Blogex renders HTML using its own<br/>templates in blogex/components.ex (post_index, post_show)
+    
+    PhoenixApp->>PhoenixApp: Render blog_html/index.html.heex (via BlogHTML)
+    
+    Note over PhoenixApp,Browser: Individual Post Request<br/>(GET /blog/engineering/:slug)
+
+    Browser->>PhoenixApp: HTTP GET /blog/engineering/new-post
+    PhoenixApp->>PhoenixApp: FirehoseWeb.BlogController.show
+    PhoenixApp->>Blogex: Get post by slug
+    Blogex->>Engineering: Read markdown file
+    Blogex->>Blogex: Parse and return %Blogex.Post{}
+    
+    Note over Blogex: Blogex renders show_page for individual posts
+    
+    PhoenixApp->>PhoenixApp: Render blog_html/show.html.heex (via BlogHTML)
+    PhoenixApp->>PhoenixApp: Apply FirehoseWeb.Layouts.app layout
+    PhoenixApp->>PhoenixApp: Wrap with FirehoseWeb.Layouts.root layout
+    
+    Note over PhoenixApp: Layout provides:<br/>- Navbar (Engineering/Releases/QWAN)<br/>- Theme toggle<br/>- Global CSS (app.css with Tailwind/daisyUI)<br/>- Footer/flash messages
+    
+    PhoenixApp->>Browser: Return full HTML page
+    
+    Browser->>Browser: Render page with app styling
+```

nono.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+nono run \
+  --profile pi \
+  --allow-cwd \
+  --allow /Users/willem/.local/share/mise \
+  --allow /Users/willem/.pi \
+  --read /Users/willem/.git \
+  --read-file /Users/willem/.gitconfig \
+  --allow /Users/willem/Library/Caches/mise \
+  --allow-net \
+  -- pi --verbose 

planner_request.md

@@ -0,0 +1,61 @@
+# Refactoring Plan for Firehose Blog Controller Tests
+
+## Context
+Based on context.md, we have a Phoenix blog controller with repetitive validation tests that need refactoring.
+
+## Goals
+1. Extract test helpers to reduce code duplication
+2. Standardize test naming conventions
+3. Reorganize tests to follow defensive programming flow
+4. Add missing negative test coverage
+5. Create separate contexts for different refactorings
+
+## Recommended Planner Agents
+
+### 1. TestHelperExtractor Agent
+**Purpose**: Handle Smell 1 (Repetitive Validation Tests) and Smell 4 (Redundant Layout Assertions)
+
+**Tasks**:
+- Extract page validation test logic into `test_page_fallback/2` helper
+- Create `assert_has_app_layout/1` helper for layout assertions
+- Move helpers to support module or test case
+
+**Context Isolation**: This can run in a separate test context without affecting controller logic.
+
+### 2. TestOrganizer Agent  
+**Purpose**: Handle Smell 5 (Test Order) and Smell 3 (Inconsistent Naming)
+
+**Tasks**:
+- Reorder test blocks: validation first, then success cases, then edge cases
+- Standardize all test descriptions to follow pattern: "GET /blog/:type/:slug returns [result]"
+- Rename describe blocks to follow semantic order
+
+**Context Isolation**: Pure test organization, no production code changes.
+
+### 3. CoverageExpander Agent
+**Purpose**: Handle Smell 2 (Missing Negative Tests) and Smell 8 (Edge Cases)
+
+**Tasks**:
+- Add test for unknown blog_id (`/blog/invalid`)
+- Add test for empty page parameter (`?page=`)
+- Add test for very large page numbers
+- Add test for invalid blog halt behavior (Smell 6)
+
+**Context Isolation**: Adds new tests without modifying existing logic.
+
+### 4. ResponseHelperCreator Agent
+**Purpose**: Handle Smell 7 (Mixed Response Types)
+
+**Tasks**:
+- Create `assert_html/2` and `assert_json/2` helpers
+- Ensure proper content-type verification
+- Update existing tests to use new helpers
+
+## Execution Strategy
+Run each agent in isolated contexts:
+1. TestHelperExtractor → creates helper functions
+2. ResponseHelperCreator → builds response assertions  
+3. TestOrganizer → reorganizes existing structure
+4. CoverageExpander → adds new test cases
+
+This keeps the main thread clean and allows focused changes per agent.

refactor_conn_aliasing.sh

@@ -0,0 +1,188 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+usage() {
+  cat <<'EOF'
+Usage: refactor_conn_aliasing.sh [OPTIONS] FILE...
+  --dry-run    Show diff without modifying files
+  --help       Show usage
+EOF
+}
+
+DRY_RUN=false
+FILES=()
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --dry-run) DRY_RUN=true; shift ;;
+    --help)    usage; exit 0 ;;
+    -*)        echo "Unknown option: $1" >&2; usage >&2; exit 1 ;;
+    *)         FILES+=("$1"); shift ;;
+  esac
+done
+
+if [[ ${#FILES[@]} -eq 0 ]]; then
+  echo "Error: no files specified" >&2
+  usage >&2
+  exit 1
+fi
+
+for file in "${FILES[@]}"; do
+  if [[ ! -f "$file" ]]; then
+    echo "Warning: $file not found, skipping" >&2
+    continue
+  fi
+
+  tmpfile=$(mktemp)
+  trap "rm -f '$tmpfile'" EXIT
+
+  awk '
+  # Detect trigger line: conn = VERB(conn, ARGS)
+  # where VERB is get/post/put/patch/delete/head/options
+  /^[[:space:]]*conn = (get|post|put|patch|delete|head|options)\(conn, / {
+    trigger_line = $0
+    # Extract leading whitespace
+    match($0, /^[[:space:]]*/)
+    indent = substr($0, RSTART, RLENGTH)
+
+    # Extract verb and args from: conn = verb(conn, args)
+    rest = $0
+    sub(/^[[:space:]]*conn = /, "", rest)
+    # rest is now: verb(conn, args)
+    paren_pos = index(rest, "(")
+    verb = substr(rest, 1, paren_pos - 1)
+    # args portion: everything after "conn, " up to the trailing ")"
+    inner = substr(rest, paren_pos + 1)
+    sub(/\)$/, "", inner)
+    # inner is: conn, args
+    sub(/^conn, /, "", inner)
+    args = inner
+
+    # Read the next non-blank line
+    triggered = 1
+    next
+  }
+
+  triggered == 1 {
+    # Skip blank lines, accumulating them
+    if ($0 ~ /^[[:space:]]*$/) {
+      blank_lines = blank_lines $0 "\n"
+      next
+    }
+
+    next_line = $0
+    triggered = 0
+
+    # Now look ahead: count how many subsequent lines (until scope boundary)
+    # reference "conn" — to decide Case 4 vs Cases 1-3
+    # We already have next_line. Check if next_line references conn.
+    # Then peek further lines.
+
+    # For simplicity: check if next_line matches Case 1, 2, or 3 patterns.
+    # If it does, check the line AFTER that for more conn references (Case 4 override).
+
+    # Case 1: var = helper(conn, status)
+    # helpers: html_response, json_response, text_response, response, redirected_to
+    case1 = 0
+    if (match(next_line, /^[[:space:]]*([a-z_]+) = (html_response|json_response|text_response|response|redirected_to)\(conn, [^)]+\)$/, m1)) {
+      case1 = 1
+      c1_var = m1[1]
+      c1_helper = m1[2]
+      # Extract status from helper(conn, status)
+      match(next_line, /\(conn, ([^)]+)\)/, m1s)
+      c1_status = m1s[1]
+    }
+
+    # Case 2: assert helper(conn, status) with optional =~ "..."
+    case2 = 0
+    if (match(next_line, /^[[:space:]]*assert (html_response|json_response|text_response|response|redirected_to)\(conn, ([^)]+)\)(.*)$/, m2)) {
+      case2 = 1
+      c2_helper = m2[1]
+      c2_status = m2[2]
+      c2_tail = m2[3]
+    }
+
+    # Case 3: assert %{...} = helper(conn, status)
+    case3 = 0
+    if (match(next_line, /^[[:space:]]*assert (%\{[^}]*\}) = (html_response|json_response|text_response|response|redirected_to)\(conn, ([^)]+)\)$/, m3)) {
+      case3 = 1
+      c3_pattern = m3[1]
+      c3_helper = m3[2]
+      c3_status = m3[3]
+    }
+
+    # If we matched Case 1, 2, or 3, emit the merged line
+    if (case1) {
+      print indent c1_var " = conn |> " verb "(" args ") |> " c1_helper "(" c1_status ")"
+      if (blank_lines != "") printf "%s", blank_lines
+      blank_lines = ""
+      next
+    }
+    if (case2) {
+      print indent "assert conn |> " verb "(" args ") |> " c2_helper "(" c2_status ")" c2_tail
+      if (blank_lines != "") printf "%s", blank_lines
+      blank_lines = ""
+      next
+    }
+    if (case3) {
+      print indent "assert " c3_pattern " = conn |> " verb "(" args ") |> " c3_helper "(" c3_status ")"
+      if (blank_lines != "") printf "%s", blank_lines
+      blank_lines = ""
+      next
+    }
+
+    # If next_line references conn at all, this is Case 4 territory
+    # (multiple uses without a recognized single-merge pattern)
+    if (next_line ~ /conn/) {
+      # Case 4: rename to response
+      print indent "response = conn |> " verb "(" args ")"
+      if (blank_lines != "") printf "%s", blank_lines
+      blank_lines = ""
+      # Replace conn with response in next_line
+      gsub(/conn/, "response", next_line)
+      print next_line
+      # Continue replacing conn->response in subsequent lines until scope boundary
+      renaming = 1
+      next
+    }
+
+    # No conn reference on next line — leave trigger unchanged (fallback)
+    print trigger_line
+    if (blank_lines != "") printf "%s", blank_lines
+    blank_lines = ""
+    print next_line
+    next
+  }
+
+  # Renaming mode for Case 4: replace conn with response until scope boundary
+  renaming == 1 {
+    # Scope boundary: blank line, "end", reduced indentation, or new conn = assignment
+    if ($0 ~ /^[[:space:]]*$/ || $0 ~ /^[[:space:]]*end$/ || $0 ~ /^[[:space:]]*conn =/) {
+      renaming = 0
+      print
+      next
+    }
+    gsub(/conn/, "response")
+    print
+    next
+  }
+
+  # Normal mode: pass through
+  {
+    if (blank_lines != "") {
+      printf "%s", blank_lines
+      blank_lines = ""
+    }
+    print
+  }
+
+  BEGIN { triggered = 0; renaming = 0; blank_lines = "" }
+  ' "$file" > "$tmpfile"
+
+  if $DRY_RUN; then
+    diff -u "$file" "$tmpfile" || true
+  else
+    mv "$tmpfile" "$file"
+    echo "Refactored: $file"
+  fi
+done

test.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+set -e
+/home/vscode/.local/bin/mise trust /workspaces/firehose/mise.toml 2>/dev/null
+eval "$(/home/vscode/.local/bin/mise activate bash)"
+cd /workspaces/firehose/app
+mix deps.get
+mix test