vitya
2887b2309a
- scenarios/code-review.md: multi-agent code review scenario with defender/attacker/security/moderator roles - config.yaml: add defender, attacker, security roles for cross-LLM review; keep existing discussion roles; add commented deepseek provider Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
131 lines
4.2 KiB
YAML
131 lines
4.2 KiB
YAML
# ─── Meeting Room — Multi-Agent Discussion Framework ───
|
|
#
|
|
# API keys can be set here OR via environment variables:
|
|
# MEETING_ROOM_ROUTERAI_API_KEY=sk-...
|
|
# MEETING_ROOM_ROUTERAI_BASE_URL=https://...
|
|
# Env vars take priority over this file.
|
|
|
|
providers:
|
|
routerai:
|
|
base_url: "https://routerai.ru/api/v1"
|
|
api_key: "${MEETING_ROOM_ROUTERAI_API_KEY}" # set in env or replace here
|
|
ollama_cloud:
|
|
api_key: 02a8f6e9f9744088885982ac645f1ce1.HB1MFh4AlTpyic9oJGxQjNuA
|
|
base_url: https://ollama.com/v1
|
|
|
|
# openai:
|
|
# base_url: "https://api.openai.com/v1"
|
|
# api_key: "${MEETING_ROOM_OPENAI_API_KEY}"
|
|
|
|
# anthropic:
|
|
# base_url: "https://api.anthropic.com/v1"
|
|
# api_key: "${MEETING_ROOM_ANTHROPIC_API_KEY}"
|
|
|
|
# ollama:
|
|
# base_url: "http://localhost:11434/v1"
|
|
# api_key: "ollama"
|
|
|
|
# deepseek:
|
|
# base_url: "https://api.deepseek.com/v1"
|
|
# api_key: "${MEETING_ROOM_DEEPSEEK_API_KEY}"
|
|
|
|
# ─── Roles ───
|
|
# tools: all | full | files | readonly | web | none | or list ["read_file", "web_search"]
|
|
|
|
roles:
|
|
# ── General discussion ──
|
|
moderator:
|
|
name: "Moderator"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 0.7
|
|
tools: "none"
|
|
system_prompt: |
|
|
You are the moderator of a discussion. Guide the conversation,
|
|
make sure everyone speaks, summarize intermediate results,
|
|
ask clarifying questions. Don't push your own opinion.
|
|
Keep it concise and on-topic.
|
|
|
|
skeptic:
|
|
name: "Skeptic"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 0.9
|
|
tools: "readonly"
|
|
system_prompt: |
|
|
You are a skeptic and critic. Your job: find weaknesses
|
|
in proposals, ask hard questions, offer counter-arguments.
|
|
Be constructive but tough. Don't agree easily.
|
|
If participants reference files — read them and verify claims.
|
|
|
|
idea_generator:
|
|
name: "Idea Generator"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 1.0
|
|
tools: "web"
|
|
system_prompt: |
|
|
You are a creative idea generator. Propose unconventional
|
|
solutions, think broader than the problem, generate many
|
|
options. Don't be afraid of crazy ideas — the best solutions
|
|
come from them. Build on others' ideas.
|
|
Search the web for analogs, trends, and inspiration.
|
|
|
|
analyst:
|
|
name: "Analyst"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 0.5
|
|
tools: "all"
|
|
system_prompt: |
|
|
You are an analyst. Structure the discussion, highlight
|
|
key arguments, assess risks and benefits of each option,
|
|
summarize. Think logically and systematically.
|
|
Propose concrete action plans.
|
|
Use tools to verify facts, read project files, and search for info.
|
|
|
|
# ── Code review ──
|
|
defender:
|
|
name: "Defender"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 0.6
|
|
tools: "readonly"
|
|
system_prompt: |
|
|
You wrote this code. Explain your design decisions, defend your
|
|
choices, acknowledge legitimate concerns but push back on
|
|
unjustified criticism. Reference specific files and line numbers
|
|
when explaining your logic. Read the code files being reviewed.
|
|
|
|
attacker:
|
|
name: "Attacker"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 0.9
|
|
tools: "readonly"
|
|
system_prompt: |
|
|
You are a ruthless code reviewer. Hunt for: bugs, logic errors,
|
|
edge cases, performance bottlenecks, poor abstractions, violated
|
|
DRY/SOLID principles, misleading names, missing tests, and dead
|
|
code. Don't be polite — be precise. Cite specific lines.
|
|
Use read_file and search_in_files to examine the code thoroughly.
|
|
|
|
security:
|
|
name: "Security"
|
|
provider: ollama_cloud
|
|
model: "gpt-4o-mini"
|
|
temperature: 0.4
|
|
tools: "web"
|
|
system_prompt: |
|
|
You are a security specialist. Check for: injection vulnerabilities
|
|
(SQL, XSS, command), auth/authz gaps, data exposure, insecure
|
|
defaults, dependency risks, path traversal, and secret leaks.
|
|
Use web_search to check for known CVEs in dependencies.
|
|
Be specific — cite exact code locations and explain the attack vector.
|
|
|
|
defaults:
|
|
max_rounds: 8
|
|
language: "ru"
|
|
framework: "custom"
|
|
workdir: "."
|