Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/selvarajmurugesan90/ops-engineering-skillsnpx agentmods add skills/selvarajmurugesan90/ops-engineering-skills/agent-bad-response-triage-and-root-cause-classificationWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/selvarajmurugesan90/ops-engineering-skills/agent-bad-response-triage-and-root-cause-classification)<a href="https://agentmods.dev/skills/selvarajmurugesan90/ops-engineering-skills/agent-bad-response-triage-and-root-cause-classification"><img src="https://agentmods.dev/badge/skills/selvarajmurugesan90/ops-engineering-skills/agent-bad-response-triage-and-root-cause-classification/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/selvarajmurugesan90/ops-engineering-skills/agent-bad-response-triage-and-root-cause-classification"><img src="https://agentmods.dev/badge/skills/selvarajmurugesan90/ops-engineering-skills/agent-bad-response-triage-and-root-cause-classification.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00146 | $0.03919 |
| Opus 5 | $0.00073 | $0.01959 |
| Sonnet 5 | $0.00029 | $0.00784 |
| Haiku 4.5 | $0.00015 | $0.00392 |
Grade A, and why
agent-bad-response-triage-and-root-cause-classification scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 300 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Bad Response Triage and Root Cause Classification
Purpose
A single reported bad response — a wrong answer, a fabricated fact, an unsafe or off-policy reply, an action that shouldn't have happened — looks identical on the surface regardless of why it happened, but the fix is completely different depending on the cause. Patching a prompt in response to what was actually a tool timeout fixes nothing and adds prompt cruft; rolling back a model version in response to what was actually a stale retrieval index wastes an incident cycle and hides the real problem. This skill is a triage runbook: reproduce the exact conditions that produced the bad response, then work through a deliberate decision tree to classify the root cause into one of five buckets — prompt issue, tool failure, retrieval issue, model behavior change, or genuine edge case — before deciding on a fix. It assumes an eval harness and guardrail layer already exist (or should); this skill is what happens between "someone reported a bad response" and "here's the regression case and the fix," which agent-evaluation-and-guardrails covers on the prevention side.
When to use
- A user, support ticket, or monitoring alert reports one specific bad, wrong, or harmful agent response and it needs root-causing before a fix is proposed.
- Deciding whether an observed failure is a one-off (genuine edge case) or a systemic issue (prompt, tool, retrieval, or model) that will recur.
- A stakeholder is pressuring for an immediate prompt patch and you need to first confirm the prompt is actually the cause.
- After a model provider version bump, a tool schema change, or a retrieval-index update, and a report comes in that might be linked to that change.
- Building or refining an incident-response runbook specifically for agent/LLM output issues, distinct from traditional application incidents.
Prerequisites & environment
- Access to the full transcript of the reported interaction: every model call (system prompt, user input, prior turns), every tool call and its raw result, and the final output — not just the final answer shown to the user. If your agent doesn't log this today, treat "add full transcript logging" as a blocking prerequisite, not optional polish.
- The ability to reproduce a call with pinned inputs: the exact prompt version, tool-schema version, model version/identifier, and (for RAG-backed agents) the retrieval index snapshot or timestamp in effect at the time of the original response.
- A changelog of recent changes to the system: prompt edits, tool schema changes, model version bumps or provider-side model updates, and ingestion/re-indexing runs — with timestamps, so they can be correlated against when the bad response occurred.
- Access to the eval suite (see agent-evaluation-and-guardrails) so a confirmed root cause can be turned into a permanent regression case.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 300 lines · 146 tokens per session scan A c7926289fcba
agent-bad-response-triage-and-root-cause-classification is a skill published in the GitHub repository selvarajmurugesan90/ops-engineering-skills (39 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 146 tokens to every session and 3,919 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
systematic-debugging
Use when debugging a failing test, build error, or runtime issue that isn't immediately obvious. Guides a 4-phase root cause analysis instead of random fix attempts.
diagnose
Trace from a reproduced symptom to the source code that causes it. Pin the specific file and approximate line, rate confidence in the cause and clarity of the fix independently, and always propose a concrete fix.
repro-admin
Reproduce an EmDash admin UI bug. Attach a container, start the demo dev server, drive the admin with agent-browser using the dev-bypass session, and capture the reproduction as screenshots plus a replayable transcript.
log-error-digest
Analyze log files to troubleshoot errors, identify peak error periods, and produce error clustering, frequency statistics, and time distribution reports. Supports JSON, syslog, and Nginx formats with automatic detection. Use when a user uploads a .log file and asks to analyze errors, find patterns, debug issues, or…
byted-util-volcengine-detect-retry
An orchestration workflow for Volcengine Cloud Detect, a service that checks websites or network endpoints from test locations.