feat(validate): catch template reference errors before runtime

[PolyphonyRequiem]

Summary

Enhanced conductor validate to catch Jinja2 template reference errors before runtime. Previously, conductor validate only checked YAML schema (Pydantic) — stale agent references, missing workflow inputs, and undeclared dependencies only surfaced at runtime after minutes of execution and real API costs.

What it catches

Errors (validation fails)

• Stale agent references: {{ old_agent_name.output.field }} where old_agent_name doesn't exist
• Missing workflow input references: {{ workflow.input.missing_field }} where missing_field isn't declared in input:
• Stale references in script args, input_mapping, system_prompt: All template-bearing fields are scanned

Warnings (validation passes with notes)

• Undeclared dependencies in explicit mode: Agent prompt references {{ a.output.val }} but doesn't declare a.output in its input: list

Example

agents:
  - name: writer
    prompt: |
      Based on {{ old_agent_name.output.findings }}
      Write about {{ workflow.input.nonexistent_field }}

$ conductor validate stale.yaml
╭─ Validation Failed ─╮
│ - agent 'writer' prompt references unknown agent 'old_agent_name'. Available: researcher, writer
│ - agent 'writer' prompt references unknown workflow input 'nonexistent_field'. Declared inputs: topic
╰──────────────────────╯

Changes

• src/conductor/config/validator.py:
  • Added _TEMPLATE_REF_PATTERN and _WORKFLOW_INPUT_REF_PATTERN for scanning templates
  • Added _collect_template_strings() to gather all Jinja2 template strings from an agent
  • Added _validate_template_references() with agent name, workflow input, and explicit mode checks
  • Updated validate_workflow_config() signature to accept optional workflow_path
• src/conductor/cli/validate.py:
  • Wired semantic validation (validate_workflow_config) into the CLI validate command — it previously only ran Pydantic schema validation
  • Displays warnings for non-fatal issues

Testing

• 280 config tests pass (no regressions)
• Tested manually with stale references, missing inputs, valid workflows, and explicit mode scenarios

[codecov-commenter]

Codecov Report

❌ Patch coverage is 83.83838% with 16 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (main@dbcf20f). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
src/conductor/config/validator.py	85.39%	13 Missing ⚠️
src/conductor/cli/validate.py	70.00%	3 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #125   +/-   ##
=======================================
  Coverage        ?   84.85%           
=======================================
  Files           ?       53           
  Lines           ?     7261           
  Branches        ?        0           
=======================================
  Hits            ?     6161           
  Misses          ?     1100           
  Partials        ?        0           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.