Add production dependency validation via validate_needs()

[Copilot]

Catch misconfigured needs chains before job submission by validating that each declared dependency will actually produce the data products required by the downstream analysis.

Pipeline metadata (pipeline.py)

• Added available_outputs and required_inputs class-level list attributes to Pipeline — declare what a pipeline produces/consumes
• Added get_actual_outputs(production) and get_actual_inputs(production) methods returning those lists by default; subclasses can override for conditional logic based on production configuration

Dependency validation (analysis.py)

• Added validate_needs() to the Analysis base class — resolves dependency names to analysis objects, then checks each required input against the outputs advertised by every dependency, issuing a UserWarning per unsatisfied requirement

# In a pipeline subclass:
class BayesWave(Pipeline):
    available_outputs = ["psd"]

class Bilby(Pipeline):
    required_inputs = ["psd"]

# At build/submission time:
bilby_analysis.validate_needs()
# UserWarning: Production 'Prod1' requires 'psd' but no dependency provides it

The method is a no-op when the pipeline declares no required_inputs, preserving backward compatibility with all existing pipelines.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Add Production Dependency Validation</issue_title>
<issue_description>Summary: Validate that productions in needs will provide required data products before submission.

Description:

Add validate_needs() method to Analysis class that:

• Checks if productions listed in needs provide the required inputs (we can find this via the enhancements described in Add Pipeline Input/Output Metadata System #94)
• Warns (or errors) if required data products won't be available
• Runs at build/submission time, not runtime

Example validation:

def validate_needs(self):
    required = self.pipeline.get_actual_inputs(self)
    for requirement in required:
        satisfied = any(
            requirement in dep.pipeline.get_actual_outputs(dep)
            for dep in self.needs
        )
        if not satisfied:
            warnings.warn(f"Production requires '{requirement}' but no dependency provides it")

Benefits:

• Catch configuration errors before wasting compute time
• Better user feedback about missing dependencies
• Foundation for smarter workflow orchestration</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Add Production Dependency Validation #95

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.