Red team analysis with interactive fix selection. Identifies issues, generates fix options, and lets you choose which fixes to apply.
Runs interactive red team analysis that identifies issues and lets you select fixes to apply. Use when you want to find vulnerabilities in your code or conversation and get actionable remediation options.
/plugin marketplace add abossenbroek/abossenbroek-claude-plugins/plugin install red-agent@abossenbroek-claude-pluginsRed team analysis with interactive fix selection. Identifies issues, generates fix options, and lets you choose which fixes to apply.
/redteam-w-fix [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 entry point for red team analysis with fix planning. Your 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 the fix-coordinator agent:
Task: Launch fix-coordinator agent
Agent: agents/fix-coordinator.md
Prompt: [Include the full YAML snapshot]
The coordinator will:
findings_with_fixesParse the coordinator's YAML output. For each finding, present a question using AskUserQuestion.
Batching Rule: AskUserQuestion supports max 4 questions per call.
For each finding, create a question:
AskUserQuestion(questions=[
{
"question": "[finding_id]: [finding_title]\nSeverity: [severity] | How should we fix this?",
"header": "[finding_id]", # Max 12 chars
"multiSelect": false,
"options": [
{
"label": "[option_a_label]",
"description": "[option_a_description - first 100 chars]"
},
{
"label": "[option_b_label]",
"description": "[option_b_description - first 100 chars]"
},
# ... up to 3 options
]
},
# ... up to 4 questions per batch
])
The "Other" option is automatically included by AskUserQuestion.
After each batch, if more findings remain, call AskUserQuestion again.
Based on user selections, generate an expert end-user summary.
Format:
# Red Team Fixes - Selected Actions
## Summary
[N] issues addressed | Touches: [list of affected components]
---
## [finding_id]: [finding_title]
**Issue**: [Brief description of the problem and its risk]
**Selected fix**: [Selected option label]
**What changes**:
- [Change 1]
- [Change 2]
- [Change 3]
**Why this over alternatives**:
[Brief explanation of why this option was selected over others]
**Watch out for**: [Any risks or things to test]
---
[Repeat for each selected fix]
---
## Suggested Order
1. [First thing to do]
2. [Second thing to do]
3. [Continue...]
If the user selected "Other" for any finding, include their custom input in the summary.
Return the implementation summary DIRECTLY to the user.
DO NOT:
ONLY return the final implementation summary.
This command is the BRIDGE between main session and red team work:
If the coordinator returns no findings:
# Red Team Analysis Complete
No issues at CRITICAL, HIGH, or MEDIUM severity were identified.
The analysis found [N] LOW/INFO level observations which don't require fixes.
Run `/redteam` for the full report if interested.
If a batch of questions returns with all "skip" or empty responses:
/redteam-w-fix standard[RF-001] [AG-002] [CM-003]
RF-001: Invalid inference in auth
Severity: HIGH | How should we fix this?
○ A: Add validation [LOW]
○ B: Refactor flow [MEDIUM]
○ C: Type-safe handlers [HIGH]
○ Other