Log Triage

Skill: Log Triage & Fix Plan (Enterprise Debugging)

Source: ETHPanda
Track: Enterprise & Team Skills → Debugging
Goal: Turn raw error logs into (1) root-cause hypotheses, (2) reproducible steps, (3) prioritized fix plan, (4) verification checklist.

When to use

Use this skill when you have:

• CI failure logs (GitHub Actions, GitLab CI, Jenkins, etc.)
• Runtime crashes, stack traces, uncaught exceptions
• Build failures (TypeScript, Python, Rust, Go, Java, Solidity/Hardhat)
• Dependency / environment / permission / network issues

Required inputs (ask user to provide)

1. The error log (full text preferred)
2. What command was running (e.g. pnpm test, hardhat test, pytest, cargo test)
3. Environment (OS, Node/Python/Rust versions, CI provider if any)
4. Recent changes (commit, dependency bumps, config edits) if known

Output format (must follow)

1. 1-paragraph summary of what failed and where
2. Error signature(s) (key lines you matched)
3. Root-cause candidates (ranked, with confidence and reasoning)
4. Fix plan (ranked, each step includes exact commands or code areas to check)
5. Verification checklist (how to confirm the fix)
6. If blocked: minimal questions to unblock (max 3)

Triage procedure

1. Extract the first fatal error (often earlier than the last line)
2. Classify into categories:
   • Syntax/type errors
   • Missing dependency / wrong version
   • Build toolchain mismatch
   • Env var / secret missing
   • Permission / path / file system
   • Network / rate limit / RPC issues
   • Test flakiness / timing
3. Match known patterns (e.g. "Cannot find module", "TypeError", "revert", "out of gas", "403/429", etc.)
4. Propose minimal reproduction locally
5. Provide safest fix first (pin version / clean install / config correction) before invasive refactors

Constraints

• Don't hallucinate files that don't exist; when unsure, say "check whether X exists".
• Prefer reversible changes (lockfile pin, config tweak) over broad refactors.
• Always include a verification step (rerun exact command / CI workflow).

Available Scripts

log_triage.py

Analyzes error logs using pattern matching and generates a structured Markdown report.

Usage:

python scripts/log_triage.py --file <log_file> --format md

Input:

• --file: Path to log file
• --format: Output format (md or txt)

Output:

• Matched error signatures
• Prioritized fix plan
• Verification checklist

Supported Error Patterns

Example Usage

Scenario 1: CI Build Failed

User Query:

"My GitHub Actions workflow failed with this log: [paste log]"

Agent Response:

## Log Triage Report

### Summary
Node.js build failed due to missing module 'viem'. The dependency exists in 
package.json but wasn't installed in CI environment.

### Matched Error Signatures
1. **node_missing_module**: `Cannot find module 'viem'`

### Fix Plan (Prioritized)
1. Run clean install: `rm -rf node_modules && pnpm install`
2. Check lockfile changes and Node version consistency (CI vs local)
3. If monorepo, ensure package is listed in the right package.json

### Verification Checklist
- [ ] Re-run the exact failing command locally
- [ ] Re-run CI workflow and confirm green
- [ ] Add a regression test if applicable

Scenario 2: Hardhat Test Failure

User Query:

"My hardhat test is failing with HH600 error"

Agent Response:

## Log Triage Report

### Summary
Hardhat compilation failed due to compiler version mismatch between 
hardhat.config and contract pragma.

### Matched Error Signatures
1. **hardhat_compile**: `HH600: Compilation failed`

### Fix Plan (Prioritized)
1. Run `pnpm hardhat compile --show-stack-traces` locally
2. Check Solidity compiler version in hardhat config matches pragma
3. If error about artifacts/cache: remove `artifacts/ cache/` and recompile

### Verification Checklist
- [ ] Contracts compile successfully
- [ ] Tests pass locally
- [ ] CI passes

Best Practices

1. Extract First Error: Often the first error causes cascading failures
2. Minimal Reproduction: Simplify to smallest failing case
3. Reversible Fixes: Try safe fixes (clean install, version pin) before refactoring
4. Verify Thoroughly: Re-run exact command + CI workflow
5. Document Root Cause: Add regression test or config comment

Integration with SpoonReactSkill

from spoon_ai.agents import SpoonReactSkill

agent = SpoonReactSkill(
    name="debug_assistant",
    skill_paths=["./enterprise-skills/debugging/log-triage"],
    scripts_enabled=True
)

await agent.activate_skill("log-triage")

result = await agent.run(
    "My CI failed with this error: [paste log]. What's wrong and how do I fix it?"
)
print(result)

Context Variables

• {{log_file}}: Path to log file
• {{command}}: Command that failed
• {{environment}}: Environment details
• {{recent_changes}}: Recent commits or dependency changes