AI
Agent Incident Postmortem
🔵 Stable🕐 updated 2026-07-01
🔷 SkillSpec L3
pm-agentops
Run a blameless postmortem for an incident caused by an AI agent or LLM feature — hallucinated facts shipped to users, runaway tool use, prompt injection, cost blowouts, or wrong actions taken autonomously. Use when asked to write up an AI incident, analyse why an agent did something wrong, or produce corrective actions after an LLM failure. Produces a structured postmortem with trace reconstruction, a root-cause layer analysis, and corrective actions including a permanent regression case. For non-AI production incidents use incident-postmortem.
What to give it
▸What the agent did — and what it should have done
▸The trace — the full request: system prompt, context, tool calls and results, output. If no trace exists, that absence is itself a finding
▸Blast radius — how many users/requests, over what window, and whether it's ongoing
▸Detection — how it was noticed (user report? monitor? luck?) and how long after it started
✅ The bar it holds itself to
Every skill in this library self-verifies — these are this skill's own quality checks, straight from its definition.
✓The postmortem is blameless toward humans *and* useful about the system — "prompt engineer error" and "model hallucinated" are both banned conclusions
✓Root cause identifies the earliest layer that could have prevented impact, not just the layer that misbehaved
✓The trace (or its absence) is in the document; findings cite it
✓Failure frequency was measured or explicitly marked unknown
✓Corrective actions include the permanent regression case and at least one detection improvement
⚠️ What it refuses to do
Do not close with "improved the prompt" as the only action — the same class of output must also be caught by a guardrail or gate next time
Do not assess frequency from one replay — nondeterministic failures hide at low temperatures and reappear at scale
Do not skip the injection question when any untrusted text (web, user docs, tickets) was in the window
Do not let "the model will be better next version" close an action item — upgrades are migrations (see model-migration-plan), not fixes
Do not write it as an outage report — the system was up; the failure was behavioural, and the doc must analyse behaviour
Install
npx pm-claude-skills add --agent claude # or codex · cursor · gemini · hermes
# or one-line MCP (every skill, any client):
claude mcp add pm-skills -- npx -y pm-claude-skills-mcpRelated skills
🔌 Embed this skill
Drop this on your blog, docs, or site — it renders a "Run this skill" card:
<div data-pm-skill="agent-incident-postmortem"></div>
<script src="https://mohitagw15856.github.io/pm-claude-skills/embed.js" async></script>💬 Discussion
Agent Incident Postmortem is one of 551 open-source professional AI agent skills — all SkillSpec L3.
Try them all in the browser · ⭐ Star on GitHub · Browse the full catalog