skills.rest
incidentfox
incidentfoxOFFICIAL
Explore all skills by this author →
WebsiteGitHub

knowledge-raptor

Search and teach a RAPTOR knowledge base of runbooks, past incidents, and service dependencies.

656|82|Updated Jan 20, 2026
Search Tags:#root cause#incident response#sre#runbooks#service dependencies#Semantic Search#Knowledge Base
One-click install
npx skills add https://github.com/incidentfox/incidentfox --skill knowledge-raptor
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: knowledge-raptor
Source: https://github.com/incidentfox/incidentfox/tree/main/sre-agent/.claude/skills/knowledge-raptor
Command: npx skills add https://github.com/incidentfox/incidentfox --skill knowledge-raptor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires httpx, and includes scripts (resource) components.

What problem does it solve?

During incident response, engineers waste time rediscovering fixes that the team already solved before. This Skill queries the RAPTOR knowledge base — curated runbooks, past incident resolutions, and service dependency graphs — so investigators can find known fixes and blast radius information before digging through Confluence.

Core Features & Use Cases

  • Incident-Aware Search: Match current symptoms against runbooks and past incidents, optionally filtered by affected service.
  • Service Dependency Graph Queries: Look up dependencies, dependents, owners, runbooks, and blast radius for any service entity.
  • Similar Incident Retrieval: Find historical incidents with matching symptoms to surface known root causes and resolutions.
  • Knowledge Teaching: Record new learnings from an investigation back into the knowledge base, with automatic duplicate and contradiction detection.
  • Use Case: When payment-gateway starts returning 503 errors, search for matching runbooks, check the blast radius of the service, find similar past incidents, and after resolving, teach the root cause back so future investigations benefit.

Quick Start

Ask the agent to search the knowledge base for runbooks and past incidents matching the symptoms "pods keep crashing with OOMKilled" on the checkout service.

Frequently Asked Questions about knowledge-raptor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search a knowledge base for incident runbooks by symptoms?▼

Run search_incident.py with a --symptoms description and optionally a --service filter. It returns matching runbooks, past incidents, and service context ranked by relevance from the RAPTOR knowledge base.

How do I find the blast radius of a service dependency?▼

Use query_graph.py with --entity set to the service name and --query-type blast_radius. The script returns direct dependents, total affected count, and the list of affected services from the dependency graph.

Should I search RAPTOR or Confluence first during an incident?▼

Start with RAPTOR because it contains structured knowledge learned from past investigations, including runbooks and incident resolutions. Fall back to Confluence for detailed architecture documentation and team wikis.

Does the RAPTOR knowledge base require authentication credentials?▼

No credentials are needed. RAPTOR runs as an internal Kubernetes service and scripts connect directly to its ClusterIP address, configured through the RAPTOR_URL environment variable.

What happens if I teach duplicate or contradictory knowledge?▼

The teach endpoint automatically detects duplicates and contradictions. The response status indicates whether the knowledge was created, merged with an existing node, flagged as a duplicate, or marked as a contradiction for review.

Why do the knowledge base scripts fail with a connection error?▼

Connection errors occur when the RAPTOR service is unreachable. Verify the service is running and that the RAPTOR_URL environment variable points to the correct address, such as the in-cluster service DNS name or localhost for development.

Related Skills

Explore More Software Engineering Skills→
badhopebadhope

monitoring

Collect Prometheus metrics, route Alertmanager alerts, and create Grafana dashboards.

Community
Advanced
AI-Foundry-CoreAI-Foundry-Core
4

incident-runbook-templates

Generate incident response runbooks from templates for structured workflows.

Official
Advanced
kettleofketchupkettleofketchup

grafana

Deploy Grafana, Prometheus, Loki, and Tempo on Kubernetes with dashboards and SSO.

Community
Advanced
zts212653zts212653
2.9k

knowledge-engineering

Guides restructuring of external project documentation into structured, AI-indexable knowledge bases.

Community
Intermediate