diff --git a/.env.example b/.env.example
index fa8ba9f5..28ef7a74 100644
--- a/.env.example
+++ b/.env.example
@@ -14,8 +14,8 @@
 # In production (docker-compose.prod.yml), this is set to http://ollama:11434 automatically.
 # OLLAMA_URL=http://localhost:11434
 
-# LLM model to use via Ollama (default: qwen2.5:14b)
-# OLLAMA_MODEL=qwen2.5:14b
+# LLM model to use via Ollama (default: qwen3.5:latest)
+# OLLAMA_MODEL=qwen3.5:latest
 
 # Enable FastAPI interactive docs at /docs and /redoc (default: false)
 # DEBUG=true
diff --git a/.gitea/workflows/tests.yml b/.gitea/workflows/tests.yml
new file mode 100644
index 00000000..81c25b25
--- /dev/null
+++ b/.gitea/workflows/tests.yml
@@ -0,0 +1,40 @@
+name: Tests
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+# Runs in timmy-time:latest — tox, ruff, and all deps are pre-installed.
+# Rebuild the image only when poetry.lock changes:
+#   docker build -t timmy-time:latest .
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Lint (ruff via tox)
+        run: tox -e lint
+
+  test:
+    runs-on: ubuntu-latest
+    needs: lint
+    steps:
+      - uses: actions/checkout@v4
+      - name: Run tests (via tox)
+        env:
+          OLLAMA_URL: "http://127.0.0.1:1"
+        run: tox -e ci
+      - name: Test summary
+        if: always()
+        run: |
+          if [ -f reports/junit.xml ]; then
+            python3 -c "
+          import xml.etree.ElementTree as ET
+          root = ET.parse('reports/junit.xml').getroot()
+          t,f,e,s = (root.attrib.get(k,'0') for k in ('tests','failures','errors','skipped'))
+          print(f'Tests: {t} | Failures: {f} | Errors: {e} | Skipped: {s}')
+          "
+          fi
diff --git a/.githooks/pre-push b/.githooks/pre-push
index 416e1439..9585f64c 100755
--- a/.githooks/pre-push
+++ b/.githooks/pre-push
@@ -1,6 +1,6 @@
 #!/usr/bin/env bash
 # Pre-push hook: runs the full CI-mirror suite before allowing a push.
-# Prevents broken builds from reaching GitHub.
+# Prevents broken builds from reaching Gitea.
 #
 # Auto-activated by `make install` via git core.hooksPath.
 
diff --git a/.gitignore b/.gitignore
index feda9983..923acb83 100644
--- a/.gitignore
+++ b/.gitignore
@@ -72,6 +72,9 @@ scripts/migrate_to_zeroclaw.py
 src/infrastructure/db_pool.py
 workspace/
 
+# Gitea Actions runner state
+.runner
+
 # macOS
 .DS_Store
 .AppleDouble
diff --git a/Dockerfile b/Dockerfile
index 0b13d656..6e98036a 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,73 +1,34 @@
-# ── Timmy Time — agent image ────────────────────────────────────────────────
+# Timmy Time — unified image (CI · dev · production)
 #
-# Serves two purposes:
-#   1. `make docker-up`    → runs the FastAPI dashboard (default CMD)
-#   2. `make docker-agent` → runs a swarm agent worker (override CMD)
+# All deps pre-installed from poetry.lock; project mounted at runtime.
 #
-# Build:  docker build -t timmy-time:latest .
-# Dash:   docker run -p 8000:8000 -v $(pwd)/data:/app/data timmy-time:latest
-# Agent:  docker run -e COORDINATOR_URL=http://dashboard:8000 \
-#                    -e AGENT_NAME=Worker-1 \
-#                    timmy-time:latest \
-#                    python -m swarm.agent_runner --agent-id w1 --name Worker-1
+#   Build: docker build -t timmy-time .
+#   CI:    act_runner mounts checkout automatically
+#   Dev:   docker run -v .:/app -p 8000:8000 timmy-time tox -e dev
+#   Run:   docker run -v .:/app -p 8000:8000 timmy-time
 
-# ── Stage 1: Builder — install deps via Poetry ──────────────────────────────
-FROM python:3.12-slim AS builder
+FROM python:3.11-slim
 
-RUN apt-get update && apt-get install -y --no-install-recommends \
-        gcc curl \
+# ── System + build prereqs ────────────────────────────────────────────────────
+RUN apt-get update \
+    && apt-get install -y --no-install-recommends \
+       gcc git bash curl fonts-dejavu-core nodejs \
     && rm -rf /var/lib/apt/lists/*
 
-WORKDIR /build
-
-# Install Poetry (only needed to resolve deps, not in runtime)
-RUN pip install --no-cache-dir poetry
-
-# Copy dependency files only (layer caching)
-COPY pyproject.toml poetry.lock ./
-
-# Install deps directly from lock file (no virtualenv, no export plugin needed)
-RUN poetry config virtualenvs.create false && \
-    poetry install --only main --extras telegram --extras discord --no-root --no-interaction
-
-# ── Stage 2: Runtime ───────────────────────────────────────────────────────
-FROM python:3.12-slim AS base
-
-RUN apt-get update && apt-get install -y --no-install-recommends \
-        curl fonts-dejavu-core \
-    && rm -rf /var/lib/apt/lists/*
+# ── Python tooling ────────────────────────────────────────────────────────────
+RUN pip install --no-cache-dir poetry tox
 
+# ── Pre-install all project deps (source mounted at runtime) ──────────────────
 WORKDIR /app
+COPY pyproject.toml poetry.lock ./
+RUN poetry config virtualenvs.create false \
+    && poetry install --with dev --no-root --no-interaction \
+    && rm pyproject.toml poetry.lock
 
-# Copy installed packages from builder
-COPY --from=builder /usr/local/lib/python3.12/site-packages /usr/local/lib/python3.12/site-packages
-COPY --from=builder /usr/local/bin /usr/local/bin
-
-# ── Application source ───────────────────────────────────────────────────────
-COPY src/ ./src/
-COPY static/ ./static/
-
-# Create data directory (mounted as a volume in production)
-RUN mkdir -p /app/data
-
-# ── Non-root user for production ─────────────────────────────────────────────
-RUN groupadd -r timmy && useradd -r -g timmy -d /app -s /sbin/nologin timmy \
-    && chown -R timmy:timmy /app
-# Ensure static/ and data/ are world-readable so bind-mounted files
-# from the macOS host remain accessible when running as the timmy user.
-RUN chmod -R o+rX /app/static /app/data
-USER timmy
-
-# ── Environment ──────────────────────────────────────────────────────────────
-ENV PYTHONPATH=/app/src
-ENV PYTHONUNBUFFERED=1
-ENV PYTHONDONTWRITEBYTECODE=1
+# ── Environment ───────────────────────────────────────────────────────────────
+ENV PYTHONPATH=/app/src \
+    PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1
 
 EXPOSE 8000
-
-# ── Healthcheck ──────────────────────────────────────────────────────────────
-HEALTHCHECK --interval=30s --timeout=5s --start-period=30s --retries=3 \
-    CMD curl -f http://localhost:8000/health || exit 1
-
-# ── Default: run the dashboard ───────────────────────────────────────────────
 CMD ["uvicorn", "dashboard.app:app", "--host", "0.0.0.0", "--port", "8000"]
diff --git a/README.md b/README.md
index e2cf678d..27bdadbd 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,5 @@
 # Timmy Time — Mission Control
 
-[![Tests](https://github.com/AlexanderWhitestone/Timmy-time-dashboard/actions/workflows/tests.yml/badge.svg)](https://github.com/AlexanderWhitestone/Timmy-time-dashboard/actions/workflows/tests.yml)
 ![Python](https://img.shields.io/badge/python-3.11+-blue)
 ![Coverage](https://img.shields.io/badge/coverage-73%25-brightgreen)
 ![License](https://img.shields.io/badge/license-MIT-green)
@@ -8,28 +7,26 @@
 A local-first, sovereign AI agent system. Talk to Timmy, watch his swarm, gate
 API access with Bitcoin Lightning — all from a browser, no cloud AI required.
 
-**[Live Docs →](https://alexanderwhitestone.github.io/Timmy-time-dashboard/)**
-
 ---
 
 ## Quick Start
 
 ```bash
-git clone https://github.com/AlexanderWhitestone/Timmy-time-dashboard.git
+git clone http://localhost:3000/rockachopa/Timmy-time-dashboard.git
 cd Timmy-time-dashboard
 make install              # create venv + install deps
 cp .env.example .env      # configure environment
 
 ollama serve              # separate terminal
-ollama pull qwen2.5:14b  # Required for reliable tool calling
+ollama pull qwen3.5:latest  # Required for reliable tool calling
 
 make dev                  # http://localhost:8000
 make test                 # no Ollama needed
 ```
 
-**Note:** qwen2.5:14b is the primary model — better reasoning and tool calling
+**Note:** qwen3.5:latest is the primary model — better reasoning and tool calling
 than llama3.1:8b-instruct while still running locally on modest hardware.
-Fallback: llama3.1:8b-instruct if qwen2.5:14b is not available.
+Fallback: llama3.1:8b-instruct if qwen3.5:latest is not available.
 llama3.2 (3B) was found to hallucinate tool output consistently in testing.
 
 ---
@@ -82,7 +79,7 @@ cp .env.example .env
 | Variable | Default | Purpose |
 |----------|---------|---------|
 | `OLLAMA_URL` | `http://localhost:11434` | Ollama host |
-| `OLLAMA_MODEL` | `qwen2.5:14b` | Primary model for reasoning and tool calling. Fallback: `llama3.1:8b-instruct` |
+| `OLLAMA_MODEL` | `qwen3.5:latest` | Primary model for reasoning and tool calling. Fallback: `llama3.1:8b-instruct` |
 | `DEBUG` | `false` | Enable `/docs` and `/redoc` |
 | `TIMMY_MODEL_BACKEND` | `ollama` | `ollama` \| `airllm` \| `auto` |
 | `AIRLLM_MODEL_SIZE` | `70b` | `8b` \| `70b` \| `405b` |
diff --git a/config/providers.yaml b/config/providers.yaml
index f0044722..20d5272b 100644
--- a/config/providers.yaml
+++ b/config/providers.yaml
@@ -113,6 +113,7 @@ fallback_chains:
   # Tool-calling models (for function calling)
   tools:
     - llama3.1:8b-instruct # Best tool use
+    - qwen3.5:latest       # Qwen 3.5 — strong tool use
     - qwen2.5:7b           # Reliable tools
     - llama3.2:3b          # Small but capable
   
diff --git a/docs/SOVEREIGN_AGI_RESEARCH.md b/docs/SOVEREIGN_AGI_RESEARCH.md
index e27bb850..a05c105e 100644
--- a/docs/SOVEREIGN_AGI_RESEARCH.md
+++ b/docs/SOVEREIGN_AGI_RESEARCH.md
@@ -172,7 +172,7 @@ support:
 ```python
 class LLMConfig(BaseModel):
     ollama_url: str = "http://localhost:11434"
-    ollama_model: str = "qwen2.5:14b"
+    ollama_model: str = "qwen3.5:latest"
     # ... all LLM settings
 
 class MemoryConfig(BaseModel):
diff --git a/poetry.lock b/poetry.lock
index c892e2c0..7c30a5e8 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -537,57 +537,6 @@ files = [
     {file = "billiard-4.2.4.tar.gz", hash = "sha256:55f542c371209e03cd5862299b74e52e4fbcba8250ba611ad94276b369b6a85f"},
 ]
 
-[[package]]
-name = "black"
-version = "26.3.0"
-description = "The uncompromising code formatter."
-optional = false
-python-versions = ">=3.10"
-groups = ["dev"]
-files = [
-    {file = "black-26.3.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:135bf8a352e35b3bfba4999c256063d8d86514654599eca7635e914a55d60ec3"},
-    {file = "black-26.3.0-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:6024a2959b6c62c311c564ce23ce0eaa977a50ed52a53f7abc83d2c9eb62b8d8"},
-    {file = "black-26.3.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:264144203ea3374542a1591b6fb317561662d074bce5d91ad6afa8d8d3e4ec3d"},
-    {file = "black-26.3.0-cp310-cp310-win_amd64.whl", hash = "sha256:1a15d1386dce3af3993bf9baeb68d3e492cbb003dae05c3ecf8530a9b75edf85"},
-    {file = "black-26.3.0-cp310-cp310-win_arm64.whl", hash = "sha256:d86a70bf048235aff62a79e229fe5d9e7809c7a05a3dd12982e7ccdc2678e096"},
-    {file = "black-26.3.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:3da07abe65732483e915ab7f9c7c50332c293056436e9519373775d62539607c"},
-    {file = "black-26.3.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:fc9fd683ccabc3dc9791b93db494d93b5c6c03b105453b76d71e5474e9dfa6e7"},
-    {file = "black-26.3.0-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:8e2c7e2c5ee09ff575869258b2c07064c952637918fc5e15f6ebd45e45eae0aa"},
-    {file = "black-26.3.0-cp311-cp311-win_amd64.whl", hash = "sha256:a849286bfc3054eaeb233b6df9056fcf969ee18bf7ecb71b0257e838a0f05e6d"},
-    {file = "black-26.3.0-cp311-cp311-win_arm64.whl", hash = "sha256:c93c83af43cda73ed8265d001214779ab245fa7a861a75b3e43828f4fb1f5657"},
-    {file = "black-26.3.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:c2b1e5eec220b419e3591a0aaa6351bd3a9c01fe6291fbaf76d84308eb7a2ede"},
-    {file = "black-26.3.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:1bab64de70bccc992432bee56cdffbe004ceeaa07352127c386faa87e81f9261"},
-    {file = "black-26.3.0-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:5b6c5f734290803b7b26493ffd734b02b72e6c90d82d45ac4d5b862b9bdf7720"},
-    {file = "black-26.3.0-cp312-cp312-win_amd64.whl", hash = "sha256:7c767396af15b54e1a6aae99ddf241ae97e589f666b1d22c4b6618282a04e4ca"},
-    {file = "black-26.3.0-cp312-cp312-win_arm64.whl", hash = "sha256:765fd6ddd00f35c55250fdc6b790c272d54ac3f44da719cc42df428269b45980"},
-    {file = "black-26.3.0-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:59754fd8f43ef457be190594c07a52c999e22cb1534dc5344bff1d46fdf1027d"},
-    {file = "black-26.3.0-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:1fd94cfee67b8d336761a0b08629a25938e4a491c440951ce517a7209c99b5ff"},
-    {file = "black-26.3.0-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:6f7b3e653a90ca1ef4e821c20f8edaee80b649c38d2532ed2e9073a9534b14a7"},
-    {file = "black-26.3.0-cp313-cp313-win_amd64.whl", hash = "sha256:f8fb9d7c2496adc83614856e1f6e55a9ce4b7ae7fc7f45b46af9189ddb493464"},
-    {file = "black-26.3.0-cp313-cp313-win_arm64.whl", hash = "sha256:e8618c1d06838f56afbcb3ffa1aa16436cec62b86b38c7b32ca86f53948ffb91"},
-    {file = "black-26.3.0-cp314-cp314-macosx_10_15_x86_64.whl", hash = "sha256:d0c6f64ead44f4369c66f1339ecf68e99b40f2e44253c257f7807c5a3ef0ca32"},
-    {file = "black-26.3.0-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:ed6f0809134e51ec4a7509e069cdfa42bf996bd0fd1df6d3146b907f36e28893"},
-    {file = "black-26.3.0-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:cc6ac0ea5dd5fa6311ca82edfa3620cba0ed0426022d10d2d5d39aedbf3e1958"},
-    {file = "black-26.3.0-cp314-cp314-win_amd64.whl", hash = "sha256:884bc0aefa96adabcba0b77b10e9775fd52d4b766e88c44dc6f41f7c82787fc8"},
-    {file = "black-26.3.0-cp314-cp314-win_arm64.whl", hash = "sha256:be3bd02aab5c4ab03703172f5530ddc8fc8b5b7bb8786230e84c9e011cee9ca1"},
-    {file = "black-26.3.0-py3-none-any.whl", hash = "sha256:e825d6b121910dff6f04d7691f826d2449327e8e71c26254c030c4f3d2311985"},
-    {file = "black-26.3.0.tar.gz", hash = "sha256:4d438dfdba1c807c6c7c63c4f15794dda0820d2222e7c4105042ac9ddfc5dd0b"},
-]
-
-[package.dependencies]
-click = ">=8.0.0"
-mypy-extensions = ">=0.4.3"
-packaging = ">=22.0"
-pathspec = ">=1.0.0"
-platformdirs = ">=2"
-pytokens = ">=0.4.0,<0.5.0"
-
-[package.extras]
-colorama = ["colorama (>=0.4.3)"]
-d = ["aiohttp (>=3.10)"]
-jupyter = ["ipython (>=7.8.0)", "tokenize-rt (>=3.2.0)"]
-uvloop = ["uvloop (>=0.15.2) ; sys_platform != \"win32\"", "winloop (>=0.5.0) ; sys_platform == \"win32\""]
-
 [[package]]
 name = "celery"
 version = "5.3.1"
@@ -885,7 +834,7 @@ version = "8.3.1"
 description = "Composable command line interface toolkit"
 optional = false
 python-versions = ">=3.10"
-groups = ["main", "dev"]
+groups = ["main"]
 files = [
     {file = "click-8.3.1-py3-none-any.whl", hash = "sha256:981153a64e25f12d547d3426c367a4857371575ee7ad18df2a6183ab0545b2a6"},
     {file = "click-8.3.1.tar.gz", hash = "sha256:12ff4785d337a1bb490bb7e9c2b1ee5da3112e94a8622f26a6c77f5d2fc6842a"},
@@ -956,11 +905,11 @@ description = "Cross-platform colored terminal text."
 optional = false
 python-versions = "!=3.0.*,!=3.1.*,!=3.2.*,!=3.3.*,!=3.4.*,!=3.5.*,!=3.6.*,>=2.7"
 groups = ["main", "dev"]
-markers = "platform_system == \"Windows\" or sys_platform == \"win32\""
 files = [
     {file = "colorama-0.4.6-py2.py3-none-any.whl", hash = "sha256:4f1d9991f5acc0ca119f9d443620b77f9d6b33703e51011c16baf57afb285fc6"},
     {file = "colorama-0.4.6.tar.gz", hash = "sha256:08695f5cb7ed6e0531a20572697297273c47b8cae5a63ffc6d6ed5c201be6e44"},
 ]
+markers = {main = "platform_system == \"Windows\" or sys_platform == \"win32\"", dev = "sys_platform == \"win32\""}
 
 [[package]]
 name = "comtypes"
@@ -1741,21 +1690,6 @@ files = [
 ]
 markers = {main = "extra == \"dev\""}
 
-[[package]]
-name = "isort"
-version = "8.0.1"
-description = "A Python utility / library to sort Python imports."
-optional = false
-python-versions = ">=3.10.0"
-groups = ["dev"]
-files = [
-    {file = "isort-8.0.1-py3-none-any.whl", hash = "sha256:28b89bc70f751b559aeca209e6120393d43fbe2490de0559662be7a9787e3d75"},
-    {file = "isort-8.0.1.tar.gz", hash = "sha256:171ac4ff559cdc060bcfff550bc8404a486fee0caab245679c2abe7cb253c78d"},
-]
-
-[package.extras]
-colors = ["colorama"]
-
 [[package]]
 name = "jinja2"
 version = "3.1.6"
@@ -2247,18 +2181,6 @@ files = [
     {file = "multidict-6.7.1.tar.gz", hash = "sha256:ec6652a1bee61c53a3e5776b6049172c53b6aaba34f18c9ad04f82712bac623d"},
 ]
 
-[[package]]
-name = "mypy-extensions"
-version = "1.1.0"
-description = "Type system extensions for programs checked with the mypy type checker."
-optional = false
-python-versions = ">=3.8"
-groups = ["dev"]
-files = [
-    {file = "mypy_extensions-1.1.0-py3-none-any.whl", hash = "sha256:1be4cccdb0f2482337c4743e60421de3a356cd97508abadd57d47403e94f5505"},
-    {file = "mypy_extensions-1.1.0.tar.gz", hash = "sha256:52e68efc3284861e772bbcd66823fde5ae21fd2fdb51c62a211403730b916558"},
-]
-
 [[package]]
 name = "networkx"
 version = "3.6"
@@ -2700,36 +2622,6 @@ files = [
     {file = "packaging-26.0.tar.gz", hash = "sha256:00243ae351a257117b6a241061796684b084ed1c516a08c48a3f7e147a9d80b4"},
 ]
 
-[[package]]
-name = "pathspec"
-version = "1.0.4"
-description = "Utility library for gitignore style pattern matching of file paths."
-optional = false
-python-versions = ">=3.9"
-groups = ["dev"]
-files = [
-    {file = "pathspec-1.0.4-py3-none-any.whl", hash = "sha256:fb6ae2fd4e7c921a165808a552060e722767cfa526f99ca5156ed2ce45a5c723"},
-    {file = "pathspec-1.0.4.tar.gz", hash = "sha256:0210e2ae8a21a9137c0d470578cb0e595af87edaa6ebf12ff176f14a02e0e645"},
-]
-
-[package.extras]
-hyperscan = ["hyperscan (>=0.7)"]
-optional = ["typing-extensions (>=4)"]
-re2 = ["google-re2 (>=1.1)"]
-tests = ["pytest (>=9)", "typing-extensions (>=4.15)"]
-
-[[package]]
-name = "platformdirs"
-version = "4.9.4"
-description = "A small Python package for determining appropriate platform-specific dirs, e.g. a `user data dir`."
-optional = false
-python-versions = ">=3.10"
-groups = ["dev"]
-files = [
-    {file = "platformdirs-4.9.4-py3-none-any.whl", hash = "sha256:68a9a4619a666ea6439f2ff250c12a853cd1cbd5158d258bd824a7df6be2f868"},
-    {file = "platformdirs-4.9.4.tar.gz", hash = "sha256:1ec356301b7dc906d83f371c8f487070e99d3ccf9e501686456394622a01a934"},
-]
-
 [[package]]
 name = "pluggy"
 version = "1.6.0"
@@ -6911,61 +6803,6 @@ rate-limiter = ["aiolimiter (>=1.1,<1.3)"]
 socks = ["httpx[socks]"]
 webhooks = ["tornado (>=6.5,<7.0)"]
 
-[[package]]
-name = "pytokens"
-version = "0.4.1"
-description = "A Fast, spec compliant Python 3.14+ tokenizer that runs on older Pythons."
-optional = false
-python-versions = ">=3.8"
-groups = ["dev"]
-files = [
-    {file = "pytokens-0.4.1-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:2a44ed93ea23415c54f3face3b65ef2b844d96aeb3455b8a69b3df6beab6acc5"},
-    {file = "pytokens-0.4.1-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:add8bf86b71a5d9fb5b89f023a80b791e04fba57960aa790cc6125f7f1d39dfe"},
-    {file = "pytokens-0.4.1-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:670d286910b531c7b7e3c0b453fd8156f250adb140146d234a82219459b9640c"},
-    {file = "pytokens-0.4.1-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:4e691d7f5186bd2842c14813f79f8884bb03f5995f0575272009982c5ac6c0f7"},
-    {file = "pytokens-0.4.1-cp310-cp310-win_amd64.whl", hash = "sha256:27b83ad28825978742beef057bfe406ad6ed524b2d28c252c5de7b4a6dd48fa2"},
-    {file = "pytokens-0.4.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:d70e77c55ae8380c91c0c18dea05951482e263982911fc7410b1ffd1dadd3440"},
-    {file = "pytokens-0.4.1-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:4a58d057208cb9075c144950d789511220b07636dd2e4708d5645d24de666bdc"},
-    {file = "pytokens-0.4.1-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:b49750419d300e2b5a3813cf229d4e5a4c728dae470bcc89867a9ad6f25a722d"},
-    {file = "pytokens-0.4.1-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:d9907d61f15bf7261d7e775bd5d7ee4d2930e04424bab1972591918497623a16"},
-    {file = "pytokens-0.4.1-cp311-cp311-win_amd64.whl", hash = "sha256:ee44d0f85b803321710f9239f335aafe16553b39106384cef8e6de40cb4ef2f6"},
-    {file = "pytokens-0.4.1-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:140709331e846b728475786df8aeb27d24f48cbcf7bcd449f8de75cae7a45083"},
-    {file = "pytokens-0.4.1-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:6d6c4268598f762bc8e91f5dbf2ab2f61f7b95bdc07953b602db879b3c8c18e1"},
-    {file = "pytokens-0.4.1-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:24afde1f53d95348b5a0eb19488661147285ca4dd7ed752bbc3e1c6242a304d1"},
-    {file = "pytokens-0.4.1-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:5ad948d085ed6c16413eb5fec6b3e02fa00dc29a2534f088d3302c47eb59adf9"},
-    {file = "pytokens-0.4.1-cp312-cp312-win_amd64.whl", hash = "sha256:3f901fe783e06e48e8cbdc82d631fca8f118333798193e026a50ce1b3757ea68"},
-    {file = "pytokens-0.4.1-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:8bdb9d0ce90cbf99c525e75a2fa415144fd570a1ba987380190e8b786bc6ef9b"},
-    {file = "pytokens-0.4.1-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:5502408cab1cb18e128570f8d598981c68a50d0cbd7c61312a90507cd3a1276f"},
-    {file = "pytokens-0.4.1-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:29d1d8fb1030af4d231789959f21821ab6325e463f0503a61d204343c9b355d1"},
-    {file = "pytokens-0.4.1-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:970b08dd6b86058b6dc07efe9e98414f5102974716232d10f32ff39701e841c4"},
-    {file = "pytokens-0.4.1-cp313-cp313-win_amd64.whl", hash = "sha256:9bd7d7f544d362576be74f9d5901a22f317efc20046efe2034dced238cbbfe78"},
-    {file = "pytokens-0.4.1-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:4a14d5f5fc78ce85e426aa159489e2d5961acf0e47575e08f35584009178e321"},
-    {file = "pytokens-0.4.1-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:97f50fd18543be72da51dd505e2ed20d2228c74e0464e4262e4899797803d7fa"},
-    {file = "pytokens-0.4.1-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:dc74c035f9bfca0255c1af77ddd2d6ae8419012805453e4b0e7513e17904545d"},
-    {file = "pytokens-0.4.1-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:f66a6bbe741bd431f6d741e617e0f39ec7257ca1f89089593479347cc4d13324"},
-    {file = "pytokens-0.4.1-cp314-cp314-win_amd64.whl", hash = "sha256:b35d7e5ad269804f6697727702da3c517bb8a5228afa450ab0fa787732055fc9"},
-    {file = "pytokens-0.4.1-cp314-cp314t-macosx_11_0_arm64.whl", hash = "sha256:8fcb9ba3709ff77e77f1c7022ff11d13553f3c30299a9fe246a166903e9091eb"},
-    {file = "pytokens-0.4.1-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:79fc6b8699564e1f9b521582c35435f1bd32dd06822322ec44afdeba666d8cb3"},
-    {file = "pytokens-0.4.1-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:d31b97b3de0f61571a124a00ffe9a81fb9939146c122c11060725bd5aea79975"},
-    {file = "pytokens-0.4.1-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:967cf6e3fd4adf7de8fc73cd3043754ae79c36475c1c11d514fc72cf5490094a"},
-    {file = "pytokens-0.4.1-cp314-cp314t-win_amd64.whl", hash = "sha256:584c80c24b078eec1e227079d56dc22ff755e0ba8654d8383b2c549107528918"},
-    {file = "pytokens-0.4.1-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:da5baeaf7116dced9c6bb76dc31ba04a2dc3695f3d9f74741d7910122b456edc"},
-    {file = "pytokens-0.4.1-cp38-cp38-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:11edda0942da80ff58c4408407616a310adecae1ddd22eef8c692fe266fa5009"},
-    {file = "pytokens-0.4.1-cp38-cp38-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:0fc71786e629cef478cbf29d7ea1923299181d0699dbe7c3c0f4a583811d9fc1"},
-    {file = "pytokens-0.4.1-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:dcafc12c30dbaf1e2af0490978352e0c4041a7cde31f4f81435c2a5e8b9cabb6"},
-    {file = "pytokens-0.4.1-cp38-cp38-win_amd64.whl", hash = "sha256:42f144f3aafa5d92bad964d471a581651e28b24434d184871bd02e3a0d956037"},
-    {file = "pytokens-0.4.1-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:34bcc734bd2f2d5fe3b34e7b3c0116bfb2397f2d9666139988e7a3eb5f7400e3"},
-    {file = "pytokens-0.4.1-cp39-cp39-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:941d4343bf27b605e9213b26bfa1c4bf197c9c599a9627eb7305b0defcfe40c1"},
-    {file = "pytokens-0.4.1-cp39-cp39-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:3ad72b851e781478366288743198101e5eb34a414f1d5627cdd585ca3b25f1db"},
-    {file = "pytokens-0.4.1-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:682fa37ff4d8e95f7df6fe6fe6a431e8ed8e788023c6bcc0f0880a12eab80ad1"},
-    {file = "pytokens-0.4.1-cp39-cp39-win_amd64.whl", hash = "sha256:30f51edd9bb7f85c748979384165601d028b84f7bd13fe14d3e065304093916a"},
-    {file = "pytokens-0.4.1-py3-none-any.whl", hash = "sha256:26cef14744a8385f35d0e095dc8b3a7583f6c953c2e3d269c7f82484bf5ad2de"},
-    {file = "pytokens-0.4.1.tar.gz", hash = "sha256:292052fe80923aae2260c073f822ceba21f3872ced9a68bb7953b348e561179a"},
-]
-
-[package.extras]
-dev = ["black", "build", "mypy", "pytest", "pytest-cov", "setuptools", "tox", "twine", "wheel"]
-
 [[package]]
 name = "pyttsx3"
 version = "2.99"
@@ -7288,6 +7125,34 @@ pygments = ">=2.13.0,<3.0.0"
 [package.extras]
 jupyter = ["ipywidgets (>=7.5.1,<9)"]
 
+[[package]]
+name = "ruff"
+version = "0.15.5"
+description = "An extremely fast Python linter and code formatter, written in Rust."
+optional = false
+python-versions = ">=3.7"
+groups = ["dev"]
+files = [
+    {file = "ruff-0.15.5-py3-none-linux_armv6l.whl", hash = "sha256:4ae44c42281f42e3b06b988e442d344a5b9b72450ff3c892e30d11b29a96a57c"},
+    {file = "ruff-0.15.5-py3-none-macosx_10_12_x86_64.whl", hash = "sha256:6edd3792d408ebcf61adabc01822da687579a1a023f297618ac27a5b51ef0080"},
+    {file = "ruff-0.15.5-py3-none-macosx_11_0_arm64.whl", hash = "sha256:89f463f7c8205a9f8dea9d658d59eff49db05f88f89cc3047fb1a02d9f344010"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:ba786a8295c6574c1116704cf0b9e6563de3432ac888d8f83685654fe528fd65"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:fd4b801e57955fe9f02b31d20375ab3a5c4415f2e5105b79fb94cf2642c91440"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:391f7c73388f3d8c11b794dbbc2959a5b5afe66642c142a6effa90b45f6f5204"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:8dc18f30302e379fe1e998548b0f5e9f4dff907f52f73ad6da419ea9c19d66c8"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:1cc6e7f90087e2d27f98dc34ed1b3ab7c8f0d273cc5431415454e22c0bd2a681"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:c1cb7169f53c1ddb06e71a9aebd7e98fc0fea936b39afb36d8e86d36ecc2636a"},
+    {file = "ruff-0.15.5-py3-none-manylinux_2_31_riscv64.whl", hash = "sha256:9b037924500a31ee17389b5c8c4d88874cc6ea8e42f12e9c61a3d754ff72f1ca"},
+    {file = "ruff-0.15.5-py3-none-musllinux_1_2_aarch64.whl", hash = "sha256:65bb414e5b4eadd95a8c1e4804f6772bbe8995889f203a01f77ddf2d790929dd"},
+    {file = "ruff-0.15.5-py3-none-musllinux_1_2_armv7l.whl", hash = "sha256:d20aa469ae3b57033519c559e9bc9cd9e782842e39be05b50e852c7c981fa01d"},
+    {file = "ruff-0.15.5-py3-none-musllinux_1_2_i686.whl", hash = "sha256:15388dd28c9161cdb8eda68993533acc870aa4e646a0a277aa166de9ad5a8752"},
+    {file = "ruff-0.15.5-py3-none-musllinux_1_2_x86_64.whl", hash = "sha256:b30da330cbd03bed0c21420b6b953158f60c74c54c5f4c1dabbdf3a57bf355d2"},
+    {file = "ruff-0.15.5-py3-none-win32.whl", hash = "sha256:732e5ee1f98ba5b3679029989a06ca39a950cced52143a0ea82a2102cb592b74"},
+    {file = "ruff-0.15.5-py3-none-win_amd64.whl", hash = "sha256:821d41c5fa9e19117616c35eaa3f4b75046ec76c65e7ae20a333e9a8696bc7fe"},
+    {file = "ruff-0.15.5-py3-none-win_arm64.whl", hash = "sha256:b498d1c60d2fe5c10c45ec3f698901065772730b411f164ae270bb6bfcc4740b"},
+    {file = "ruff-0.15.5.tar.gz", hash = "sha256:7c3601d3b6d76dce18c5c824fc8d06f4eef33d6df0c21ec7799510cde0f159a2"},
+]
+
 [[package]]
 name = "safetensors"
 version = "0.7.0"
@@ -8613,4 +8478,4 @@ voice = ["pyttsx3"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.11,<4"
-content-hash = "fd2362ee582a0dcffca740b971ef13f1ddefd7dc6323d4c6a066b7ab16b7acff"
+content-hash = "bb8088a38625a65b8f7d296f912b0c1437b12c53f6b26698a91f030e82b1bf57"
diff --git a/pyproject.toml b/pyproject.toml
index 9ced87e7..0d450c05 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -9,8 +9,8 @@ description = "Mission Control for sovereign AI agents"
 readme = "README.md"
 license = "MIT"
 authors = ["Alexander Whitestone"]
-homepage = "https://alexanderwhitestone.github.io/Timmy-time-dashboard/"
-repository = "https://github.com/AlexanderWhitestone/Timmy-time-dashboard"
+homepage = "http://localhost:3000/rockachopa/Timmy-time-dashboard"
+repository = "http://localhost:3000/rockachopa/Timmy-time-dashboard"
 packages = [
     { include = "config.py", from = "src" },
     { include = "brain", from = "src" },
diff --git a/src/config.py b/src/config.py
index 72dca524..c4a6662c 100644
--- a/src/config.py
+++ b/src/config.py
@@ -14,11 +14,11 @@ class Settings(BaseSettings):
     ollama_url: str = "http://localhost:11434"
 
     # LLM model passed to Agno/Ollama — override with OLLAMA_MODEL
-    # qwen2.5:14b is the primary model — better reasoning and tool calling
+    # qwen3.5:latest is the primary model — better reasoning and tool calling
     # than llama3.1:8b-instruct while still running locally on modest hardware.
-    # Fallback: llama3.1:8b-instruct if qwen2.5:14b not available.
+    # Fallback: llama3.1:8b-instruct if qwen3.5:latest not available.
     # llama3.2 (3B) hallucinated tool output consistently in testing.
-    ollama_model: str = "qwen2.5:14b"
+    ollama_model: str = "qwen3.5:latest"
 
     # Set DEBUG=true to enable /docs and /redoc (disabled by default)
     debug: bool = False
@@ -108,7 +108,7 @@ class Settings(BaseSettings):
     # Default is False (telemetry disabled) to align with sovereign AI vision.
     telemetry_enabled: bool = False
 
-    # CORS allowed origins for the web chat interface (GitHub Pages, etc.)
+    # CORS allowed origins for the web chat interface (Gitea Pages, etc.)
     # Set CORS_ORIGINS as a comma-separated list, e.g. "http://localhost:3000,https://example.com"
     cors_origins: list[str] = ["*"]
 
@@ -302,8 +302,8 @@ if not settings.repo_root:
 
 # ── Model fallback configuration ────────────────────────────────────────────
 # Primary model for reliable tool calling (llama3.1:8b-instruct)
-# Fallback if primary not available: qwen2.5:14b
-OLLAMA_MODEL_PRIMARY: str = "qwen2.5:14b"
+# Fallback if primary not available: qwen3.5:latest
+OLLAMA_MODEL_PRIMARY: str = "qwen3.5:latest"
 OLLAMA_MODEL_FALLBACK: str = "llama3.1:8b-instruct"
 
 
diff --git a/src/dashboard/routes/swarm.py b/src/dashboard/routes/swarm.py
index 7cb9a437..741bbafc 100644
--- a/src/dashboard/routes/swarm.py
+++ b/src/dashboard/routes/swarm.py
@@ -73,19 +73,20 @@ async def swarm_live(request: Request):
 @router.websocket("/live")
 async def swarm_ws(websocket: WebSocket):
     """WebSocket endpoint for live swarm updates."""
-    await ws_manager.connect(websocket)
+    await websocket.accept()
+    # Send initial state before joining broadcast pool to avoid race conditions
+    await websocket.send_json(
+        {
+            "type": "initial_state",
+            "data": {
+                "agents": {"total": 0, "active": 0, "list": []},
+                "tasks": {"active": 0},
+                "auctions": {"list": []},
+            },
+        }
+    )
+    await ws_manager.connect(websocket, accept=False)
     try:
-        # Send initial state so frontend can clear loading placeholders
-        await websocket.send_json(
-            {
-                "type": "initial_state",
-                "data": {
-                    "agents": {"total": 0, "active": 0, "list": []},
-                    "tasks": {"active": 0},
-                    "auctions": {"list": []},
-                },
-            }
-        )
         while True:
             await websocket.receive_text()
     except WebSocketDisconnect:
diff --git a/src/infrastructure/models/multimodal.py b/src/infrastructure/models/multimodal.py
index ca543b3c..b26b77ec 100644
--- a/src/infrastructure/models/multimodal.py
+++ b/src/infrastructure/models/multimodal.py
@@ -93,6 +93,18 @@ KNOWN_MODEL_CAPABILITIES: dict[str, set[ModelCapability]] = {
         ModelCapability.VISION,
     },
     # Qwen series
+    "qwen3.5": {
+        ModelCapability.TEXT,
+        ModelCapability.TOOLS,
+        ModelCapability.JSON,
+        ModelCapability.STREAMING,
+    },
+    "qwen3.5:latest": {
+        ModelCapability.TEXT,
+        ModelCapability.TOOLS,
+        ModelCapability.JSON,
+        ModelCapability.STREAMING,
+    },
     "qwen2.5": {
         ModelCapability.TEXT,
         ModelCapability.TOOLS,
@@ -259,6 +271,7 @@ DEFAULT_FALLBACK_CHAINS: dict[ModelCapability, list[str]] = {
     ],
     ModelCapability.TOOLS: [
         "llama3.1:8b-instruct",  # Best tool use
+        "qwen3.5:latest",  # Qwen 3.5 — strong tool use
         "llama3.2:3b",  # Smaller but capable
         "qwen2.5:7b",  # Reliable fallback
     ],
diff --git a/src/infrastructure/ws_manager/handler.py b/src/infrastructure/ws_manager/handler.py
index 36c8cc1b..9beb0bfc 100644
--- a/src/infrastructure/ws_manager/handler.py
+++ b/src/infrastructure/ws_manager/handler.py
@@ -36,9 +36,15 @@ class WebSocketManager:
         self._connections: list[WebSocket] = []
         self._event_history: collections.deque[WSEvent] = collections.deque(maxlen=100)
 
-    async def connect(self, websocket: WebSocket) -> None:
-        """Accept a new WebSocket connection."""
-        await websocket.accept()
+    async def connect(self, websocket: WebSocket, *, accept: bool = True) -> None:
+        """Accept a new WebSocket connection and add it to the broadcast pool.
+
+        Args:
+            websocket: The WebSocket to register.
+            accept: If False, skip the accept() call (caller already accepted).
+        """
+        if accept:
+            await websocket.accept()
         self._connections.append(websocket)
         logger.info(
             "WebSocket connected — %d active connections",
diff --git a/src/timmy/agent.py b/src/timmy/agent.py
index 035e96d9..90475b78 100644
--- a/src/timmy/agent.py
+++ b/src/timmy/agent.py
@@ -33,6 +33,7 @@ logger = logging.getLogger(__name__)
 DEFAULT_MODEL_FALLBACKS = [
     "llama3.1:8b-instruct",
     "llama3.1",
+    "qwen3.5:latest",
     "qwen2.5:14b",
     "qwen2.5:7b",
     "llama3.2:3b",
diff --git a/src/timmy/tools.py b/src/timmy/tools.py
index 8e4fd3f2..6573e9e4 100644
--- a/src/timmy/tools.py
+++ b/src/timmy/tools.py
@@ -142,6 +142,35 @@ def calculator(expression: str) -> str:
         return f"Error evaluating '{expression}': {e}"
 
 
+def _make_smart_read_file(file_tools: FileTools) -> Callable:
+    """Wrap FileTools.read_file so directories auto-list their contents.
+
+    When the user (or the LLM) passes a directory path to read_file,
+    the raw Agno implementation throws an IsADirectoryError.  This
+    wrapper detects that case, lists the directory entries, and returns
+    a helpful message so the model can pick the right file on its own.
+    """
+    original_read = file_tools.read_file
+
+    def smart_read_file(file_name: str, encoding: str = "utf-8") -> str:
+        """Reads the contents of the file `file_name` and returns the contents if successful."""
+        # Resolve the path the same way FileTools does
+        _safe, resolved = file_tools.check_escape(file_name)
+        if _safe and resolved.is_dir():
+            entries = sorted(p.name for p in resolved.iterdir() if not p.name.startswith("."))
+            listing = "\n".join(f"  - {e}" for e in entries) if entries else "  (empty directory)"
+            return (
+                f"'{file_name}' is a directory, not a file. "
+                f"Files inside:\n{listing}\n\n"
+                "Please call read_file with one of the files listed above."
+            )
+        return original_read(file_name, encoding=encoding)
+
+    # Preserve the original docstring for Agno tool schema generation
+    smart_read_file.__doc__ = original_read.__doc__
+    return smart_read_file
+
+
 def create_research_tools(base_dir: str | Path | None = None):
     """Create tools for the research agent (Echo).
 
@@ -161,7 +190,7 @@ def create_research_tools(base_dir: str | Path | None = None):
 
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
     return toolkit
@@ -189,7 +218,7 @@ def create_code_tools(base_dir: str | Path | None = None):
 
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.save_file, name="write_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
@@ -210,12 +239,12 @@ def create_aider_tool(base_path: Path):
         def __init__(self, base_dir: Path):
             self.base_dir = base_dir
 
-        def run_aider(self, prompt: str, model: str = "qwen2.5:14b") -> str:
+        def run_aider(self, prompt: str, model: str = "qwen3.5:latest") -> str:
             """Run Aider to generate code changes.
 
             Args:
                 prompt: What you want Aider to do (e.g., "add a fibonacci function")
-                model: Ollama model to use (default: qwen2.5:14b)
+                model: Ollama model to use (default: qwen3.5:latest)
 
             Returns:
                 Aider's response with the code changes made
@@ -269,7 +298,7 @@ def create_data_tools(base_dir: str | Path | None = None):
 
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
     # Web search for finding datasets
@@ -292,7 +321,7 @@ def create_writing_tools(base_dir: str | Path | None = None):
     # File operations
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.save_file, name="write_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
@@ -320,7 +349,7 @@ def create_security_tools(base_dir: str | Path | None = None):
     # File reading for logs/configs
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
     return toolkit
@@ -342,7 +371,7 @@ def create_devops_tools(base_dir: str | Path | None = None):
     # File operations for config management
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.save_file, name="write_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
@@ -444,7 +473,7 @@ def create_full_toolkit(base_dir: str | Path | None = None):
 
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.save_file, name="write_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
@@ -586,7 +615,7 @@ def create_experiment_tools(base_dir: str | Path | None = None):
 
     base_path = Path(base_dir) if base_dir else Path(settings.repo_root)
     file_tools = FileTools(base_dir=base_path)
-    toolkit.register(file_tools.read_file, name="read_file")
+    toolkit.register(_make_smart_read_file(file_tools), name="read_file")
     toolkit.register(file_tools.save_file, name="write_file")
     toolkit.register(file_tools.list_files, name="list_files")
 
@@ -706,7 +735,7 @@ def get_all_available_tools() -> dict[str, dict]:
         },
         "aider": {
             "name": "Aider AI Assistant",
-            "description": "Local AI coding assistant using Ollama (qwen2.5:14b or deepseek-coder)",
+            "description": "Local AI coding assistant using Ollama (qwen3.5:latest or deepseek-coder)",
             "available_in": ["forge", "orchestrator"],
         },
         "prepare_experiment": {
diff --git a/tests/dashboard/test_dashboard.py b/tests/dashboard/test_dashboard.py
index 44214139..d077d100 100644
--- a/tests/dashboard/test_dashboard.py
+++ b/tests/dashboard/test_dashboard.py
@@ -82,10 +82,12 @@ def test_agents_list(client):
 
 
 def test_agents_list_metadata(client):
+    from config import settings
+
     response = client.get("/agents")
     agent = next(a for a in response.json()["agents"] if a["id"] == "default")
     assert agent["name"] == "Agent"
-    assert agent["model"] == "qwen2.5:14b"
+    assert agent["model"] == settings.ollama_model
     assert agent["type"] == "local"
 
 
diff --git a/tests/dashboard/test_mobile_scenarios.py b/tests/dashboard/test_mobile_scenarios.py
index 08e30680..7a72b4bb 100644
--- a/tests/dashboard/test_mobile_scenarios.py
+++ b/tests/dashboard/test_mobile_scenarios.py
@@ -283,15 +283,18 @@ def test_M604_airllm_print_response_delegates_to_run():
 
 def test_M605_health_status_passes_model_to_template(client):
     """Health status partial must receive the configured model name, not a hardcoded string."""
+    from config import settings
+
     with patch(
         "dashboard.routes.health.check_ollama",
         new_callable=AsyncMock,
         return_value=True,
     ):
         response = client.get("/health/status")
-    # The default model is qwen2.5:14b — it should appear from settings
+    # Model name should come from settings, not be hardcoded
     assert response.status_code == 200
-    assert "qwen2.5" in response.text  # rendered via template variable, not hardcoded literal
+    model_short = settings.ollama_model.split(":")[0]
+    assert model_short in response.text
 
 
 # ── M7xx — XSS prevention ─────────────────────────────────────────────────────
diff --git a/tests/dashboard/test_round4_fixes.py b/tests/dashboard/test_round4_fixes.py
index dd387474..e21b338e 100644
--- a/tests/dashboard/test_round4_fixes.py
+++ b/tests/dashboard/test_round4_fixes.py
@@ -7,6 +7,8 @@ and Ollama timeout parameter.
 
 from unittest.mock import MagicMock, patch
 
+import pytest
+
 # ---------------------------------------------------------------------------
 # Fix 1: /calm no longer returns 500
 # ---------------------------------------------------------------------------
@@ -103,11 +105,12 @@ def test_swarm_live_websocket_sends_initial_state(client):
 
     with client.websocket_connect("/swarm/live") as ws:
         data = ws.receive_json()
-        assert data["type"] == "initial_state"
-        assert "agents" in data["data"]
-        assert "tasks" in data["data"]
-        assert "auctions" in data["data"]
-        assert data["data"]["agents"]["total"] >= 0
+        # First message should be initial_state with swarm data
+        assert data.get("type") == "initial_state", f"Unexpected WS message: {data}"
+        payload = data.get("data", {})
+        assert "agents" in payload
+        assert "tasks" in payload
+        assert "auctions" in payload
 
 
 # ---------------------------------------------------------------------------
@@ -251,8 +254,9 @@ def test_task_full_lifecycle(client):
 # ---------------------------------------------------------------------------
 
 
+@pytest.mark.skip_ci
 def test_all_dashboard_pages_return_200(client):
-    """Smoke test: all main dashboard routes return 200."""
+    """Smoke test: all main dashboard routes return 200 (needs Ollama for /thinking)."""
     pages = [
         "/",
         "/tasks",
diff --git a/tests/e2e/test_ollama_integration.py b/tests/e2e/test_ollama_integration.py
index 833f7d09..6e8120d2 100644
--- a/tests/e2e/test_ollama_integration.py
+++ b/tests/e2e/test_ollama_integration.py
@@ -66,6 +66,7 @@ async def test_timmy_agent_with_available_model():
         pytest.skip(f"Timmy agent creation failed: {e}")
 
 
+@pytest.mark.ollama
 @pytest.mark.asyncio
 async def test_timmy_chat_with_simple_query():
     """Test that Timmy can respond to a simple chat query."""
diff --git a/tests/integrations/test_chat_bridge.py b/tests/integrations/test_chat_bridge.py
index f1f18aba..3aa0897b 100644
--- a/tests/integrations/test_chat_bridge.py
+++ b/tests/integrations/test_chat_bridge.py
@@ -240,11 +240,15 @@ class TestInviteParser:
     @pytest.mark.asyncio
     async def test_parse_image_no_deps(self):
         """parse_image returns None when pyzbar/Pillow are not installed."""
+        from unittest.mock import AsyncMock, patch
+
         from integrations.chat_bridge.invite_parser import InviteParser
 
         parser = InviteParser()
-        # With mocked pyzbar, this should gracefully return None
-        result = await parser.parse_image(b"fake-image-bytes")
+        # Mock out the Ollama vision call so we don't make a real HTTP request
+        with patch.object(parser, "_try_ollama_vision", new_callable=AsyncMock, return_value=None):
+            # With mocked pyzbar + mocked vision, this should gracefully return None
+            result = await parser.parse_image(b"fake-image-bytes")
         assert result is None
 
 
diff --git a/tests/test_smoke.py b/tests/test_smoke.py
index 4bba87d9..ae87eb12 100644
--- a/tests/test_smoke.py
+++ b/tests/test_smoke.py
@@ -6,6 +6,8 @@ crashes.  They run fast (no Ollama needed) and should stay green on
 every commit.
 """
 
+from unittest.mock import MagicMock, patch
+
 import pytest
 from fastapi.testclient import TestClient
 
@@ -14,8 +16,14 @@ from fastapi.testclient import TestClient
 def client():
     from dashboard.app import app
 
-    with TestClient(app, raise_server_exceptions=False) as c:
-        yield c
+    # Block all LLM calls so smoke tests never hit a real Ollama
+    mock_run = MagicMock()
+    mock_run.content = "Smoke test — no LLM."
+    mock_agent = MagicMock()
+    mock_agent.run.return_value = mock_run
+    with patch("timmy.agent.create_timmy", return_value=mock_agent):
+        with TestClient(app, raise_server_exceptions=False) as c:
+            yield c
 
 
 # ---------------------------------------------------------------------------
diff --git a/tests/timmy/test_agent.py b/tests/timmy/test_agent.py
index bdb0a0ee..d55661ab 100644
--- a/tests/timmy/test_agent.py
+++ b/tests/timmy/test_agent.py
@@ -51,7 +51,9 @@ def test_create_timmy_passes_ollama_url_to_model():
 
         kwargs = MockOllama.call_args.kwargs
         assert "host" in kwargs, "Ollama() must receive host= parameter"
-        assert kwargs["host"] == "http://localhost:11434"  # default from config
+        from config import settings
+
+        assert kwargs["host"] == settings.ollama_url
 
 
 def test_create_timmy_respects_custom_ollama_url():
diff --git a/tox.ini b/tox.ini
index d02e6a54..dab85e45 100644
--- a/tox.ini
+++ b/tox.ini
@@ -4,9 +4,12 @@ no_package = true
 
 # ── Base ─────────────────────────────────────────────────────────────────────
 [testenv]
-allowlist_externals = timeout, perl, docker, mkdir, bash, grep
+allowlist_externals = timeout, perl, docker, mkdir, bash, grep, ruff, pytest, mypy, uvicorn
+sitepackages = true
 commands_pre = pip install -e ".[dev]" --quiet
 
+passenv = OLLAMA_URL
+
 setenv =
     TIMMY_TEST_MODE = 1
     TIMMY_DISABLE_CSRF = 1
@@ -84,7 +87,7 @@ commands =
 # ── CI / Coverage ────────────────────────────────────────────────────────────
 
 [testenv:ci]
-description = CI test suite with coverage + JUnit XML (mirrors GitHub Actions)
+description = CI test suite with coverage + JUnit XML (mirrors Gitea Actions)
 commands =
     mkdir -p reports
     pytest tests/ \
@@ -94,7 +97,7 @@ commands =
         --cov-fail-under=73 \
         --junitxml=reports/junit.xml \
         -p no:xdist \
-        -m "not ollama and not docker and not selenium and not external_api"
+        -m "not ollama and not docker and not selenium and not external_api and not skip_ci"
 
 [testenv:coverage]
 description = Full coverage report (terminal + XML)
@@ -121,7 +124,7 @@ commands =
 # ── Pre-push (mirrors CI exactly) ────────────────────────────────────────────
 
 [testenv:pre-push]
-description = Local gate — lint + full CI suite (same as GitHub Actions)
+description = Local gate — lint + full CI suite (same as Gitea Actions)
 deps =
     ruff>=0.8.0
 commands =
@@ -136,7 +139,7 @@ commands =
         --cov-fail-under=73 \
         --junitxml=reports/junit.xml \
         -p no:xdist \
-        -m "not ollama and not docker and not selenium and not external_api"
+        -m "not ollama and not docker and not selenium and not external_api and not skip_ci"
 
 # ── Pre-commit (fast local gate) ────────────────────────────────────────────