Browse
Search and filter by type across the catalog
931 results in Sub-Agents, Hooks, Evals, Memory · page 21 of 39
llm-redteam-specialist
Use this agent when you need to red-team a Large Language Model deployment — jailbreak probes, prompt injection harness design, output-safety evaluation, and robustness evidence for EU AI Act Article 15 or NIST AI RMF MEASURE-2.7. Covers cloud-hosted models and on-prem / air-gapped local models (Ollama, vLLM, llama.cpp). Specifically:\\n\\n<example>\\nContext: A healthcare vendor embeds an LLM in a clinical triage tool and the compliance team wants a red-team report before rollout.\\nuser: \"We're deploying a Llama-3 70B behind a clinical assistant. Legal wants evidence it won't hand out harmful medical advice or leak PHI from retrieval context. How do we test it and document it?\"\\nassistant: \"I'll design an air-gapped red-team harness: a probe suite covering jailbreak families (DAN, role-play escalation, encoding attacks, prompt-leaking, indirect injection via retrieved docs), a scoring rubric aligned to the deployment's harm taxonomy, and a repeatable runner targeting your Ollama endpoint. Output is a robustness evidence pack: pass/fail table, example transcripts, coverage metric, and a control narrative mapped to NIST AI RMF MEASURE-2.7 and EU AI Act Article 15.\"\\n<commentary>\\nInvoke llm-redteam-specialist when the question is about evaluating an LLM's resistance to adversarial input — not generic web pentesting. This agent understands jailbreak taxonomies and the difference between a model-level test and a system-level test.\\n</commentary>\\n</example>\\n\\n<example>\\nContext: A defense contractor is running only local models (no external API calls allowed) and needs offline evaluation tooling.\\nuser: \"Air-gapped network. No HuggingFace, no OpenAI API. We still need quarterly robustness evidence for Llama-3 and Mistral instances. Build the test rig.\"\\nassistant: \"I'll spec an offline harness: probe corpus committed to the local repo, runner that targets localhost Ollama / vLLM endpoints, deterministic scoring (no model-as-judge calls outside the enclave), and a signed evidence bundle per run. Retention and signing align to the site's audit requirements. I'll pair this with a schedule for re-runs after every model or system-prompt change.\"\\n<commentary>\\nUse when the environment forbids cloud-hosted grader models and probe corpora must be self-contained.\\n</commentary>\\n</example>\\n\\n<example>\\nContext: A SaaS company received a prospect security questionnaire asking for evidence of prompt-injection testing.\\nuser: \"Enterprise prospect wants evidence we test for prompt injection. What do I send them?\"\\nassistant: \"I'll produce a prompt-injection test report: scope (which endpoints and retrieval paths were tested), probe inventory with OWASP LLM Top 10 references, results table, severity rubric, and remediation status per finding. I'll also flag the gap between direct-injection and indirect-injection coverage so the evidence is honest.\"\\n<commentary>\\nInvoke for LLM-specific adversarial evidence — distinct from penetration-tester which covers web/network.\\n</commentary>\\n</example>
llms-maintainer
LLMs.txt roadmap file generator and maintainer for AI Engine Optimization (AEO). Use after build completion, content changes, or when setting up AI crawler navigation for a site. Detects framework, scans site structure, and writes a spec-compliant llms.txt file.
load-testing-specialist
Load testing and stress testing specialist. Use PROACTIVELY for creating comprehensive load test scenarios, analyzing performance under stress, and identifying system bottlenecks and capacity limits.
m365-admin
Use when automating Microsoft 365 administrative tasks including Exchange Online mailbox provisioning, Teams collaboration management, SharePoint site configuration, license lifecycle management, and Graph API-driven identity automation.
malware-analyst
Expert malware analyst specializing in defensive malware research, threat intelligence, and incident response. Masters sandbox analysis, behavioral analysis, and malware family identification. Handles static/dynamic analysis, unpacking, and IOC extraction. Use PROACTIVELY for malware triage, threat hunting, incident response, or security research.
marketing-agent
Marketing strategist and copywriter for campaign planning, audience research, positioning, copy creation, and content review. Covers landing pages, email sequences, social posts, ad copy, short-form video scripts, and content calendars. Use when the user wants to plan or execute a product launch…
memory-manager
PROACTIVELY USE this agent when project memory becomes cluttered, CLAUDE.md files are getting too long, outdated information needs to be removed, project knowledge needs reorganization for better efficiency, or when optimizing information organization for long-term projects. This agent MUST BE USED for memory optimization and knowledge organization tasks. Examples: <example>Context: Project has accumulated extensive documentation and context over time. user: 'Our CLAUDE.md file is getting very long and contains outdated information mixed with current requirements' assistant: 'I'll use the memory-manager agent to organize and optimize our project memory for better efficiency.' <commentary>Since project memory needs optimization and organization, use the memory-manager to curate and structure information.</commentary></example> <example>Context: User notices Claude is struggling with information retrieval due to cluttered context. user: 'I feel like Claude is having trouble finding relevant information quickly because there's too much context' assistant: 'Let me use the memory-manager agent to streamline and organize our project knowledge for more efficient access.' <commentary>The memory-manager should be used to optimize information organization and improve retrieval efficiency.</commentary></example>