test(#296): fix TopicSeedData assertion + seed script docs

.github/workflows/README.md

@@ -60,7 +60,7 @@ The GitHub Actions user needs these permissions:
 - Lambda full access
 - DynamoDB full access
 - API Gateway full access
-- S3 access for SAM artifacts
+- S3 access for deployment artifacts (e.g. Lambda bundles, Pulumi state backups as configured)
 - Secrets Manager read access
 
 ## Deployment Flow

.gitignore

@@ -406,6 +406,7 @@ security-reports/
 vulnerability-reports/
 .snyk
 .artifacts/*
+.artifacts-downloaded/
 !.artifacts/security/
 .artifacts/security/*
 !.artifacts/security/.gitignore

coaching/pyproject.toml

@@ -67,8 +67,8 @@ dependencies = [
 [project.optional-dependencies]
 dev = [
     # Testing
-    "pytest>=7.4.0,<9.0.0",
-    "pytest-asyncio>=0.21.0,<1.0.0",
+    "pytest>=9.0.3,<10.0.0",
+    "pytest-asyncio>=1.3.0,<2.0.0",
     "pytest-mock>=3.11.1,<4.0.0",
     "pytest-cov>=4.1.0,<6.0.0",
     "pytest-xdist>=3.3.1,<4.0.0",
@@ -109,8 +109,8 @@ packages = ["src"]
 
 [tool.uv]
 dev-dependencies = [
-    "pytest>=7.4.0",
-    "pytest-asyncio>=0.21.0",
+    "pytest>=9.0.3",
+    "pytest-asyncio>=1.3.0",
     "pytest-mock>=3.11.1",
     "black>=23.7.0",
     "ruff>=0.1.0",

coaching/requirements-dev.txt

@@ -1,17 +1,17 @@
--r requirements.txt
-
-# Additional runtime extras
-email-validator==2.1.0.post1
-python-multipart==0.0.20
-
-# Development tools
-pytest==9.0.3
-pytest-asyncio==1.3.0
-pytest-mock==3.14.0
-pytest-cov==6.0.0
-black==23.7.0
-ruff==0.8.4
-pylint==2.17.4
-mypy==1.4.1
-types-PyYAML
-types-requests
+-r requirements.txt
+
+# Additional runtime extras
+email-validator==2.1.0.post1
+python-multipart==0.0.20
+
+# Development tools
+pytest==9.0.3
+pytest-asyncio==1.3.0
+pytest-mock==3.14.0
+pytest-cov==6.0.0
+black==23.7.0
+ruff==0.8.4
+pylint==2.17.4
+mypy==1.4.1
+types-PyYAML
+types-requests

coaching/src/scripts/README.md

@@ -1,105 +1,125 @@
-# Coaching Topics Seed Scripts
-
-Scripts for seeding coaching topics to the DynamoDB topics table.
-
-## Overview
-
-These scripts seed coaching topics into the `purposepath-topics-{env}` DynamoDB table.
-
-## Scripts
-
-### `seed_topics.py`
-Main seeding script that:
-- Creates all coaching topics in DynamoDB from the endpoint registry
-- Supports 44 topics total
-- Is idempotent (safe to run multiple times)
-- Can update existing topics or skip them
-
-## Usage
-
-```bash
-# Set environment variables
-export AWS_PROFILE=purposepath-dev
-export AWS_REGION=us-east-1
-export STAGE=dev
-
-# Run seed script
-cd coaching
-python -m src.scripts.seed_topics
-
-# Run with options
-python -m src.scripts.seed_topics --force-update  # Update existing
-python -m src.scripts.seed_topics --dry-run       # Preview changes
-python -m src.scripts.seed_topics --validate-only # Validate only
-```
-
-## Prerequisites
-
-1. **Infrastructure Deployed**: DynamoDB topics table must exist
-2. **AWS Credentials**: Configure AWS profile or credentials
-3. **Dependencies**: Install Python dependencies
-- Old YAML files remain untouched (cleaned up later)
-
-## Idempotency
-
-The seed script is **idempotent**:
-- ✅ Safe to run multiple times
-- ✅ Skips existing topics
-- ✅ Does not overwrite data
-- ✅ Reports what was created vs skipped
-
-## Troubleshooting
-
-### "Topic already exists, skipping"
-This is normal if you've run the script before. The script won't overwrite existing topics.
-
-### "No existing prompts found for migration"
-This means no old YAML file exists in S3. Admin must create prompts via the API endpoints.
-
-### DynamoDB Access Denied
-Ensure your AWS credentials have DynamoDB PutItem permission for the table.
-
-### S3 Access Denied
-Ensure your AWS credentials have S3 GetObject and PutObject permissions for the bucket.
-
-## Verification
-
-After seeding, verify with:
-
-```bash
-# Check DynamoDB table
-aws dynamodb scan \
-  --table-name purposepath-llm-prompts-dev \
-  --index-name topic_type-index \
-  --expression-attribute-values '{":type":{"S":"conversation_coaching"}}' \
-  --filter-expression "topic_type = :type"
-
-# Check S3 prompts
-aws s3 ls s3://purposepath-coaching-prompts-dev/prompts/ --recursive
-
-# Or use verification script
-./scripts/verify_topics.sh dev
-```
-
-## Next Steps
-
-After seeding:
-1. **Verify**: Run verification script
-2. **Test**: Test conversation endpoints work with new topics
-3. **Create Prompts**: If no prompts were migrated, create them via admin API
-4. **Deploy**: Topics are now ready for use
-
-## Related Issues
-
-- #78: Infrastructure setup
-- #79: Domain models and repository
-- #80: Admin API endpoints
-- #81: Service layer migration
-- #82: This seed script (final epic task)
-
-## Notes
-
-- Seed script should be run **once per environment** after infrastructure deployment
-- The CoachingTopic enum remains for backward compatibility
-- Old YAML prompts can be removed in a future cleanup issue
-- Prompts can be updated via the admin API after seeding
+# Coaching Topics Seed Scripts
+
+Scripts for seeding coaching topics to the DynamoDB topics table.
+
+## Overview
+
+These scripts seed coaching topics into the `purposepath-topics-{env}` DynamoDB table.
+
+## Scripts
+
+### `seed_topics.py`
+Main seeding script that:
+- Creates all coaching topics in DynamoDB from the endpoint registry
+- Seeds every topic defined in the endpoint registry / topic seed data
+- Is idempotent (safe to run multiple times)
+- Can update existing topics or skip them
+
+## Usage
+
+The script imports `coaching.src.*`. Set **`PYTHONPATH` to the repository root** (the parent folder of the `coaching` package), then run the module as `coaching.src.scripts.seed_topics`.
+
+```bash
+# Set environment variables
+export AWS_PROFILE=purposepath-dev
+export AWS_REGION=us-east-1
+export STAGE=dev
+
+# Repository root = PurposePath_AI (folder that contains the `coaching` directory)
+export PYTHONPATH="$(pwd)"
+
+# Run seed script (from repo root or from coaching — see below)
+cd coaching
+uv run python -m coaching.src.scripts.seed_topics
+
+# Run with options
+uv run python -m coaching.src.scripts.seed_topics --force-update  # Update existing
+uv run python -m coaching.src.scripts.seed_topics --dry-run       # Preview changes
+uv run python -m coaching.src.scripts.seed_topics --validate-only # Validate only
+
+# Seed a single topic (example: strategy_alignment_evaluation, issue #296)
+uv run python -m coaching.src.scripts.seed_topics --topic-id strategy_alignment_evaluation --force-update
+```
+
+**PowerShell (Windows), from repository root:**
+
+```powershell
+$env:AWS_PROFILE = "purposepath-dev"
+$env:AWS_REGION = "us-east-1"
+$env:STAGE = "dev"
+$env:PYTHONPATH = "C:\path\to\PurposePath_AI"
+cd coaching
+uv run python -m coaching.src.scripts.seed_topics --topic-id strategy_alignment_evaluation --force-update
+```
+
+Legacy path `python -m src.scripts.seed_topics` only works if your environment exposes the same import layout as CI (see `pyproject.toml` `pythonpath` for tests).
+
+## Prerequisites
+
+1. **Infrastructure Deployed**: DynamoDB topics table must exist
+2. **AWS Credentials**: Configure AWS profile or credentials
+3. **Dependencies**: Install Python dependencies (`uv sync` in `coaching/`)
+
+## Idempotency
+
+The seed script is **idempotent**:
+- ✅ Safe to run multiple times
+- ✅ Skips existing topics
+- ✅ Does not overwrite data
+- ✅ Reports what was created vs skipped
+
+## Troubleshooting
+
+### "Topic already exists, skipping"
+This is normal if you've run the script before. The script won't overwrite existing topics.
+
+### "No existing prompts found for migration"
+This means no old YAML file exists in S3. Admin must create prompts via the API endpoints.
+
+### DynamoDB Access Denied
+Ensure your AWS credentials have DynamoDB PutItem permission for the table.
+
+### S3 Access Denied
+Ensure your AWS credentials have S3 GetObject and PutObject permissions for the bucket.
+
+## Verification
+
+After seeding, verify with:
+
+```bash
+# Check DynamoDB table
+aws dynamodb scan \
+  --table-name purposepath-llm-prompts-dev \
+  --index-name topic_type-index \
+  --expression-attribute-values '{":type":{"S":"conversation_coaching"}}' \
+  --filter-expression "topic_type = :type"
+
+# Check S3 prompts
+aws s3 ls s3://purposepath-coaching-prompts-dev/prompts/ --recursive
+
+# Or use verification script
+./scripts/verify_topics.sh dev
+```
+
+## Next Steps
+
+After seeding:
+1. **Verify**: Run verification script
+2. **Test**: Test conversation endpoints work with new topics
+3. **Create Prompts**: If no prompts were migrated, create them via admin API
+4. **Deploy**: Topics are now ready for use
+
+## Related Issues
+
+- #78: Infrastructure setup
+- #79: Domain models and repository
+- #80: Admin API endpoints
+- #81: Service layer migration
+- #82: This seed script (final epic task)
+
+## Notes
+
+- Seed script should be run **once per environment** after infrastructure deployment
+- The CoachingTopic enum remains for backward compatibility
+- Old YAML prompts can be removed in a future cleanup issue
+- Prompts can be updated via the admin API after seeding

coaching/src/scripts/seed_topics.py

@@ -15,6 +15,7 @@
     --deactivate-orphans   Deactivate topics that no longer have endpoints
 
 Examples:
+    # Set PYTHONPATH to the repository root (parent of the `coaching` package), then:
     # Seed all new topics (skip existing)
     python -m coaching.src.scripts.seed_topics
 
@@ -32,6 +33,8 @@
 
     # Deactivate orphaned topics
     python -m coaching.src.scripts.seed_topics --deactivate-orphans
+
+    See coaching/src/scripts/README.md for PowerShell / uv invocations.
 """
 
 import argparse

coaching/tests/integration/test_unified_ai_engine.py

@@ -117,7 +117,8 @@ def test_all_topics_have_required_fields(self):
             assert topic_data.topic_name
             assert topic_data.topic_type in ["single_shot", "conversation_coaching"]
             assert topic_data.category
-            assert topic_data.model_code
+            assert topic_data.basic_model_code
+            assert topic_data.premium_model_code
             assert 0.0 <= topic_data.temperature <= 2.0
             assert topic_data.max_tokens > 0