feat: add redthread ECC bundle

.agents/skills/redthread/SKILL.md

@@ -0,0 +1,117 @@
+```markdown
+# redthread Development Patterns
+
+> Auto-generated skill from repository analysis
+
+## Overview
+This skill teaches the core development patterns and workflows used in the `redthread` Python codebase. You'll learn the project's coding conventions, how to add new research modules or phases, manage research dependencies, and follow the repository's commit and testing standards. This guide is designed to help contributors maintain consistency and quality across the codebase.
+
+## Coding Conventions
+
+### File Naming
+- Use **snake_case** for all Python files.
+  - Example: `gepa_phase.py`, `test_gepa_adapter.py`
+
+### Import Style
+- Use **relative imports** within modules.
+  - Example:
+    ```python
+    from .models import GEPAResult
+    from . import gepa_adapter
+    ```
+
+### Export Style
+- Use **named exports** (explicitly define what is exported from each module).
+  - Example:
+    ```python
+    __all__ = ["GEPAResult", "run_gepa_phase"]
+    ```
+
+### Commit Messages
+- Follow **conventional commit** style.
+  - Prefixes: `feat`, `docs`
+  - Example: `feat: add GEPA phase 0 module and tests`
+
+## Workflows
+
+### Add New Research Module
+**Trigger:** When adding a new research phase, module, or major algorithmic component (e.g., GEPA phase, Pareto selection, adapter) to the research pipeline.  
+**Command:** `/new-research-module`
+
+1. **Create or modify Python modules** under `src/redthread/research/`.
+   - Example: `src/redthread/research/gepa_phase.py`
+2. **Update or add test files** under `tests/`.
+   - Example: `tests/test_gepa_phase0.py`
+3. **Update supporting files** as needed to integrate the new module.
+   - Example: `src/redthread/research/models.py`, `workspace.py`
+4. **Ensure code quality and correctness:**
+   - Run `ruff` for linting.
+   - Run `mypy` for type checking.
+   - Run all relevant tests and ensure they pass.
+
+**Example:**
+```python
+# src/redthread/research/gepa_phase.py
+from .models import GEPAResult
+
+def run_gepa_phase(data):
+    # Implementation here
+    pass
+
+__all__ = ["run_gepa_phase"]
+```
+
+```python
+# tests/test_gepa_phase0.py
+from src.redthread.research.gepa_phase import run_gepa_phase
+
+def test_gepa_phase_basic():
+    result = run_gepa_phase(sample_data)
+    assert result is not None
+```
+
+---
+
+### Add or Update Research Dependency
+**Trigger:** When introducing or updating a research-related dependency (e.g., gepa, litellm), especially for new research phases or scripts.  
+**Command:** `/add-research-dependency`
+
+1. **Edit `pyproject.toml`** to add or update the dependency under an `[optional]` group.
+   - Example:
+     ```toml
+     [project.optional-dependencies]
+     research-gepa = ["gepa>=1.2.0"]
+     ```
+2. **Update `uv.lock`** to reflect the new or updated dependency.
+3. **Import the dependency lazily or optionally** in the relevant research module.
+   - Example:
+     ```python
+     try:
+         import gepa
+     except ImportError:
+         gepa = None
+     ```
+4. **Verify that core installs do not require the optional dependency.**
+   - Ensure the main package can be installed without the research dependency unless explicitly requested.
+
+---
+
+## Testing Patterns
+
+- **Test files** are named with the pattern `test_*.py` and located in the `tests/` directory.
+  - Example: `tests/test_gepa_adapter.py`
+- **Testing framework** is not explicitly specified, but tests follow standard Python test conventions.
+- **Example test:**
+  ```python
+  def test_gepa_adapter_runs():
+      from src.redthread.research.gepa_adapter import run_adapter
+      assert run_adapter(sample_input) == expected_output
+  ```
+
+## Commands
+
+| Command                | Purpose                                                        |
+|------------------------|----------------------------------------------------------------|
+| /new-research-module   | Scaffold and integrate a new research module or phase          |
+| /add-research-dependency | Add or update an optional research dependency for research   |
+```

.agents/skills/redthread/agents/openai.yaml

@@ -0,0 +1,6 @@
+interface:
+  display_name: "Redthread"
+  short_description: "Repo-specific patterns and workflows for redthread"
+  default_prompt: "Use the redthread repo skill to follow existing architecture, testing, and workflow conventions."
+policy:
+  allow_implicit_invocation: true

.claude/commands/add-new-research-module.md

@@ -0,0 +1,37 @@
+---
+name: add-new-research-module
+description: Workflow command scaffold for add-new-research-module in redthread.
+allowed_tools: ["Bash", "Read", "Write", "Grep", "Glob"]
+---
+
+# /add-new-research-module
+
+Use this workflow when working on **add-new-research-module** in `redthread`.
+
+## Goal
+
+Implements a new research module or phase, including code, tests, and plumbing for new GEPA phases or components.
+
+## Common Files
+
+- `src/redthread/research/*.py`
+- `tests/test_*.py`
+
+## Suggested Sequence
+
+1. Understand the current state and failure mode before editing.
+2. Make the smallest coherent change that satisfies the workflow goal.
+3. Run the most relevant verification for touched files.
+4. Summarize what changed and what still needs review.
+
+## Typical Commit Signals
+
+- Create or modify one or more new Python modules under src/redthread/research/ (e.g., gepa_phase.py, gepa_adapter.py, gepa_pareto.py, etc.)
+- Update or add corresponding test files under tests/ (e.g., test_gepa_phase0.py, test_gepa_adapter.py, test_gepa_pareto.py)
+- Update or create any necessary supporting files (e.g., models.py, workspace.py) to integrate the new module
+- Ensure ruff and mypy pass, and all relevant tests are green
+
+## Notes
+
+- Treat this as a scaffold, not a hard-coded script.
+- Update the command if the workflow evolves materially.

.claude/commands/add-or-update-research-dependency.md

@@ -0,0 +1,38 @@
+---
+name: add-or-update-research-dependency
+description: Workflow command scaffold for add-or-update-research-dependency in redthread.
+allowed_tools: ["Bash", "Read", "Write", "Grep", "Glob"]
+---
+
+# /add-or-update-research-dependency
+
+Use this workflow when working on **add-or-update-research-dependency** in `redthread`.
+
+## Goal
+
+Adds or updates an optional research dependency (e.g., gepa, litellm) and ensures it is properly pinned and grouped in pyproject.toml and lockfile.
+
+## Common Files
+
+- `pyproject.toml`
+- `uv.lock`
+- `src/redthread/research/*.py`
+
+## Suggested Sequence
+
+1. Understand the current state and failure mode before editing.
+2. Make the smallest coherent change that satisfies the workflow goal.
+3. Run the most relevant verification for touched files.
+4. Summarize what changed and what still needs review.
+
+## Typical Commit Signals
+
+- Edit pyproject.toml to add or update the dependency under an [optional] group (e.g., [research-gepa])
+- Update uv.lock to reflect the new/updated dependency
+- Ensure that the dependency is imported lazily or optionally in the relevant research module
+- Verify that core installs do not require the optional dependency
+
+## Notes
+
+- Treat this as a scaffold, not a hard-coded script.
+- Update the command if the workflow evolves materially.