Experiment Provenance Ledger

Experiment Provenance Ledger

Mission

Preserve experiment provenance across code, data, environment, parameters, metrics, artifacts, and human decisions.

Current pain point: Teams cannot explain which model, data, code, or parameters produced a result after the fact.

Why this skill exists: A provenance ledger creates an immutable audit trail independent of any single tracking tool.

Operating Rules

• Start by restating the decision this skill is supporting.
• Inspect local artifacts first before asking for context.
• Treat scripts as evidence collectors, not as substitutes for judgment.
• Preserve raw data, notebooks, configs, and model artifacts unless the user explicitly asks for mutation.
• Mark missing context as unknown, not provided, or owner decision; do not invent it.
• Classify each blocker as stop, fix-first, monitor, accepted risk, or owner decision.
• Use the output contract exactly unless the user asks for a different format.
• Keep final recommendations auditable: every important claim needs evidence, assumption, or caveat.

What This Skill Must Do

• Transform a vague request into an explicit Experiment Provenance Ledger decision with named owner, evidence, assumptions, and action threshold.
• Inspect local artifacts first: data extracts, schemas, notebooks, SQL, pipeline configs, model reports, tickets, and prior decisions.
• Separate mechanical checks from expert judgment so another reviewer can reproduce what was checked and what was inferred.
• Classify findings as stop, fix-first, monitor, accepted risk, or owner decision instead of producing generic advice.
• Preserve raw evidence and never silently mutate data, notebooks, model artifacts, or production configs.
• Return a decision artifact that can be handed to a data scientist, ML engineer, governance reviewer, or stakeholder without hidden context.

Required Inputs

• repo path
• dataset identifiers
• environment files
• training command
• metrics
• artifact paths
• decision notes

If an input is missing, inspect available files first. Ask only for information that cannot be recovered from the workspace and would change the recommendation.

Workflow

1. Capture code version, dirty state, environment, data references, parameters, seeds, and runtime.
2. Record metrics and artifacts with owner, purpose, and retention expectation.
3. Link human decisions, exclusions, and model-selection rationale.
4. Map entries to MLflow, DVC, W&B, or plain JSON depending on the stack.
5. Return a ledger and missing-provenance risk list.

Red Flags

• unversioned dataset
• dirty code used for final metric
• missing seed
• manual artifact rename
• model chosen without logged rationale

When a red flag appears, slow down and surface it in Risks. A red flag does not always mean stop, but it must change the recommendation or the confidence level.

Resources

• Read references/playbook.md for the skill-specific checklist, scoring rubric, and failure modes.
• Read references/acceptance-tests.md before forward-testing clean, messy, and adversarial requests.
• Read references/agent-portability.md when adapting this skill to Claude, Gemini, Copilot, Cursor, Windsurf, Gravity, LangGraph, CrewAI, AutoGen, or local agents.
• Read references/quality-rubric.md when reviewing whether the output meets senior data-science standards.
• Use scripts/quick_validate_skill.py . --strict before publishing or installing the skill.

Use only the specific reference file needed for the task; keep context small.

Output Contract

Return these sections unless the user requests another format:

• Problem: the decision, artifact, model, data source, or workflow being handled.
• Inputs: files, data, stakeholder context, assumptions, and missing context used.
• Checks performed: concrete inspections, scripts, and reasoning checks.
• Findings: prioritized observations with evidence.
• Decision: go, fix-first, stop, proceed-with-risk, or needs-owner-review.
• Risks: unresolved blockers, caveats, and owner decisions.
• Next actions: smallest useful follow-up steps.
• Artifacts: generated files, specs, reports, scripts, or links.

Final Checks

• Did the response answer the actual decision, not just analyze data?
• Did it preserve evidence and avoid silent mutation?
• Did it name unknowns and owner decisions?
• Did it include at least one concrete next action?
• Did it avoid overclaiming causality, readiness, compliance, or production safety?