Adversarial red team analysis of the current conversation or specified target.
Launch an adversarial analysis to find vulnerabilities in your conversation, code, or files. Use when you want to stress-test AI responses for security flaws, reasoning errors, or manipulation risks.
/plugin marketplace add abossenbroek/abossenbroek-claude-plugins/plugin install red-agent@abossenbroek-claude-pluginsAdversarial red team analysis of the current conversation or specified target.
/redteam [mode] [target]
mode (optional):
quick - Fast 2-3 vector analysis, skip groundingstandard - Balanced 5-6 vectors with basic grounding (default)deep - All categories + meta-analysis with full groundingfocus:[category] - Deep dive on specific category (e.g., focus:reasoning-flaws)target (optional):
conversation - Current conversation context (default)file:path - Analyze specific filecode - Analyze recent git changesYou are the MINIMAL entry point for red team analysis. Your ONLY job is to:
Launch the pal-availability-checker agent to detect if PAL MCP is available:
Task: Launch pal-availability-checker agent
Agent: agents/pal-availability-checker.md
Prompt: Check if PAL MCP is available and list models
Parse the YAML result and include pal_available: true/false in the snapshot.
This step is NON-BLOCKING - continue regardless of result. PAL is optional.
Determine mode and target from the command arguments:
standardconversationCreate a YAML-formatted snapshot of the current session. DO NOT include raw conversation - structure it as data:
snapshot:
mode: [parsed mode]
target: [parsed target]
pal_available: [true/false from Step 1]
pal_models: [list of models if available, empty if not]
conversational_arc:
message_count: [count of messages in conversation]
phases:
- phase: "[phase name]"
messages: [range]
summary: "[what happened in this phase]"
key_transitions:
- from_msg: [number]
to_msg: [number]
note: "[what changed and why]"
early_assumptions_carried_forward:
- assumption: "[assumption text]"
introduced_at: [message number]
still_active: [true/false]
claims:
- id: C[N]
text: "[factual claim made by assistant]"
speaker: assistant
confidence: [stated_as_fact|hedged|uncertain]
message_num: [source message number]
files_read:
- path: [file path]
summary: "[brief description of content/purpose]"
tools_invoked:
- tool: [tool name]
command: "[command or action]"
outcome: "[result summary]"
decisions:
- decision: "[decision made]"
rationale: "[stated reason]"
assumptions_explicit:
- "[explicitly stated assumption]"
Use the Task tool to launch a SINGLE agent:
Task: Launch red-team-coordinator agent
Agent: agents/red-team-coordinator.md
Prompt: [Include the full YAML snapshot]
Return the coordinator's markdown report DIRECTLY to the user.
DO NOT:
ONLY return the final sanitized markdown report.
This command is the FIREWALL between main session and red team work: