- Home
- Skills
- Agents & Orchestration
- Training Pipeline Failure Forensics
Works with the AI Agents you already use
Training Pipeline Failure Forensics
Rank evidence-backed causes of ML training crashes or silent degradation, then design the cheapest discriminating test and minimal reproduction.
$19.99
Training Pipeline Failure Forensics
Example session with this skill installed
A nightly training job still completes, but validation MAE increased from about 6.2 to 9.8 after a warehouse schema migration. Join-step row counts doubled for one day, dependencies are loosely pinned, and the last code merge was three weeks earlier. Build an evidence-ranked RCA and the smallest test that separates the leading causes.
- Read your context and instructions
- Compiled the training pipeline failure
ARTIFACT: TRAINING-FAILURE-RCA
MODE: FORENSICS
SURFACE: SILENT DEGRADATION
-
Data contract / leakage — Highly likely
Evidence: one-day join fan-out aligned with the schema migration; duplicates may persist in the rolling training window.
Cheapest test: count duplicate trip IDs in the affected join output. -
Dependency drift — Weak signal
Evidence: loose version pins keep it possible; timing is weaker.
Cheapest test: diff the last-good and current environment locks.
MINIMAL REPRODUCTION
Check the affected partition, deduplicate the window, retrain once with the same seed, and compare MAE.
BOUNDARY
The cause remains unconfirmed until a discriminating test passes.
Connects securely to your tools. The creator never sees your data.
About this skill
Turn a failed training run into a ranked investigation
Training failures appear in two forms: loud crashes and silent degradation. This skill separates the symptom from the cause, timelines what changed since the last good run, and tests six failure classes against supplied evidence.
What it produces
- A loud-versus-silent failure classification
- A change-set timeline across data, code, dependencies, infrastructure, configuration, and orchestration
- Ranked hypotheses with evidence for and against each one
- A confidence label and cheapest discriminating test per hypothesis
- A minimal reproduction path for the leading hypothesis
- A prevention line that can become a pipeline gate
Failure classes covered
Data-contract breaks, leakage, environment or dependency drift, resource and numerical failures, code defects, and orchestration faults.
Best for
- Training jobs that crash with partial logs
- Runs that complete but produce worse metrics
- Teams comparing the last good run with a failed one
- Reviewers who need a minimal test before authorizing a rerun
Boundaries
The skill ranks hypotheses; it does not prove causality, execute SQL or training jobs, repair code, access infrastructure, or guarantee that the leading hypothesis is correct. Confirmation requires the named discriminating test.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Ranks hypotheses from supplied evidence and does not prove causality, execute diagnostic commands, access infrastructure, rerun training, repair code, or guarantee recovery. Missing logs and artifacts remain explicit unknowns. Runtime routing was not executed during package validation.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 3 months ago
Permissions
No special permissions declared or detected