mirror of
https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git
synced 2026-06-19 15:59:37 +00:00
503bf2e5d7
AI Layer: - llm_router.py: routes cheap/mid/high models, enforces daily budget, caches - token_counter.py: estimates tokens, truncates to budget - response_cache.py: in-memory cache with TTL per agent - prompt_registry.py: versioned prompts with stable prefix for caching - ai_budget.yaml: model costs, agent budgets, daily limits (10 SAR/day) Guardrails: - output_validator.py: blocks fake claims + prohibited actions - cost_guard.py: prevents runaway spending Observability: - trace.py: trace_id, cost, latency, steps per pipeline run Tests: ALL PASS - 30/30 evals (100%) — 9 sectors, 30 companies - 10/10 prohibited actions blocked - 4/4 allowed actions verified - 3/3 forbidden claims blocked - 3/3 message quality checks passed https://claude.ai/code/session_01W1rJthWDkasijTdXCfxVHs
72 lines
1.8 KiB
YAML
72 lines
1.8 KiB
YAML
models:
|
|
cheap:
|
|
id: "groq/llama-3.1-8b-instant"
|
|
max_tokens: 500
|
|
cost_per_1k_input: 0.0001
|
|
cost_per_1k_output: 0.0002
|
|
use_for: ["csv_cleanup", "classification", "initial_scoring", "reply_classification"]
|
|
mid:
|
|
id: "groq/llama-3.3-70b-versatile"
|
|
max_tokens: 800
|
|
cost_per_1k_input: 0.0006
|
|
cost_per_1k_output: 0.0008
|
|
use_for: ["company_research", "website_summary", "enrichment", "content_draft"]
|
|
high:
|
|
id: "anthropic/claude-sonnet-4-20250514"
|
|
max_tokens: 1200
|
|
cost_per_1k_input: 0.003
|
|
cost_per_1k_output: 0.015
|
|
use_for: ["sales_message", "negotiation", "partnership_strategy", "eval_judge", "complex_analysis"]
|
|
|
|
agent_budgets:
|
|
company_research_agent:
|
|
model_tier: mid
|
|
max_output_tokens: 700
|
|
cache_ttl_hours: 168
|
|
scoring_agent:
|
|
model_tier: cheap
|
|
max_output_tokens: 200
|
|
cache_ttl_hours: 24
|
|
channel_strategy_agent:
|
|
model_tier: cheap
|
|
max_output_tokens: 300
|
|
cache_ttl_hours: 24
|
|
message_generation_agent:
|
|
model_tier: high
|
|
max_output_tokens: 500
|
|
cache_ttl_hours: 0
|
|
negotiation_agent:
|
|
model_tier: high
|
|
max_output_tokens: 800
|
|
cache_ttl_hours: 0
|
|
compliance_agent:
|
|
model_tier: cheap
|
|
max_output_tokens: 200
|
|
cache_ttl_hours: 720
|
|
partnership_strategist_agent:
|
|
model_tier: mid
|
|
max_output_tokens: 500
|
|
cache_ttl_hours: 168
|
|
icp_strategist_agent:
|
|
model_tier: mid
|
|
max_output_tokens: 600
|
|
cache_ttl_hours: 168
|
|
learning_agent:
|
|
model_tier: mid
|
|
max_output_tokens: 400
|
|
cache_ttl_hours: 0
|
|
content_strategy_agent:
|
|
model_tier: mid
|
|
max_output_tokens: 400
|
|
cache_ttl_hours: 24
|
|
|
|
daily_budget:
|
|
max_cost_sar: 10.0
|
|
max_requests: 500
|
|
alert_at_percent: 80
|
|
|
|
cache:
|
|
enabled: true
|
|
backend: "memory"
|
|
default_ttl_hours: 24
|