Trajectory Review
Skill Verified ActivePost-hoc diagnosis of a failed agent trajectory. Classifies the first unrecoverable step into one of nine failure categories (plan adherence, hallucinated information, invalid tool call, misread tool output, intent–plan mismatch, under-specified intent, unsupported intent, guardrail trigger, system failure) and produces an evidence-backed root-cause report.
To provide a systematic and evidence-based diagnosis of failed agent runs, enabling users to understand the root cause of errors and improve agent reliability.
Features
- Classifies failures into nine categories
- Identifies the first unrecoverable step
- Provides evidence-backed root-cause reports
- Normalizes trajectories for analysis
- Synthesizes tool contracts and goal checklists
Use Cases
- Diagnosing why an agent run produced an incorrect or incomplete result.
- Investigating user-reported agent failures and performing postmortems.
- Characterizing failure modes after model upgrades or new capability releases.
- Reviewing agent trajectories when a user requests an explanation for a failure.
Non-Goals
- Evaluating passing agent runs.
- Rerunning or debugging trajectories automatically.
- Rewriting agent prompts or skills directly; provides recommendations only.
- Generalizing findings from a single trajectory to system-wide claims without further analysis.
Trust
- info:Issues AttentionIn the last 90 days, 10 issues were opened and 48 were closed, indicating active maintenance but a closure rate below 50%.
Installation
npx skills add majiayu000/vibeguardRuns the Vercel skills CLI (skills.sh) via npx — needs Node.js locally and at least one installed skills-compatible agent (Claude Code, Cursor, Codex, …). Assumes the repo follows the agentskills.io format.
Quality Score
VerifiedSimilar Extensions
Safe Debug
99Trusted-lane debug skill for deep learning research work. Use when the user pastes a traceback, terminal error, CUDA OOM, checkpoint load failure, shape mismatch, NaN loss symptom, or training failure and wants conservative diagnosis before any patching. Do not use for broad refactoring, speculative adaptation, automatic exploratory patching, or general repository familiarization.
Node Connect
100Diagnose OpenClaw Android, iOS, or macOS node pairing, QR/setup code, route, auth, and connection failures.
Openclaw Debugging
100Debug OpenClaw model, provider, tool-surface, code-mode, streaming, and live/Crabbox behavior by choosing the right logs, probes, and proof path before changing code.
Systematic Debugging
100Systematic debugging methodology emphasizing root cause analysis over quick fixes
Troubleshooting
100Uses Chrome DevTools MCP and documentation to troubleshoot connection and target issues. Trigger this skill when list_pages, new_page, or navigate_page fail, or when the server initialization fails.
Root Cause Tracing
99Use when errors occur deep in execution and you need to trace back to find the original trigger - systematically traces bugs backward through call stack, adding instrumentation when needed, to identify source of invalid data or incorrect behavior