lightrun-error-remediation-automation
Guide deterministic runtime investigations in environments using Lightrun MCP tools, with preflight gating, recovery/resume rules, evidence-first diagnosis, PR-first fix proposal delivery, and local source-code fallback only when PR creation is not possible.
60
41%
Does it follow best practices?
Impact
89%
1.27xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/lightrun-error-remediation-automation/SKILL.mdEvaluation results
85%
26%Runtime Investigation: Intermittent Order Processing Failures
State storage discovery and reproduction-required handoff
Storage discovery documented
70%
100%
No local file for state
0%
100%
Preflight gate documented
20%
100%
Source target justified
87%
50%
Hypotheses before tools
100%
100%
Exact MCP tool identifiers
50%
50%
Single status check per action
50%
90%
Reproduction-required handoff format
40%
50%
Action cleanup documented
70%
90%
No premature fix delivery
100%
100%
State storage check precedes actions
83%
100%
88%
13%Runtime Investigation: API Rate Limiter Applying Wrong Tier
PR-first fix delivery and complete final handoff
Persistent state read first
100%
100%
Snapshot results retrieved
90%
70%
PR creation checked first
16%
83%
PR-first delivery or documented blocker
66%
91%
Local edits limited to fix files
100%
100%
Action cleanup after diagnosis
90%
70%
Cleanup status persisted
12%
62%
Diagnosis with runtime traceability
100%
100%
Final handoff completeness
83%
100%
No new action without new hypothesis
100%
100%
96%
18%Runtime Investigation: Background Job Processor Hanging Under Load
Action error mitigation and no-hit response
Hypotheses before tool selection
100%
100%
Preflight documented
42%
100%
Exact tool identifiers per hypothesis
87%
100%
Action error mitigation: troubleshooting guide consulted
16%
100%
No-hit window adjustment
90%
80%
Source re-check after no-hit
100%
100%
Reproduction-required issued for no-hit
90%
100%
No static-code-only diagnosis
66%
91%
Action cleanup both actions
100%
100%
Mitigation logged in handoff
87%
87%
Single check per action
100%
100%
- Repository
- lightrun-platform/lightrun-ai
- Commit
890166b
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.