Caching¶

Cache LLM responses to reduce costs and improve speed.

graph LR
    subgraph "Caching Flow"
        A[Request] --> B{Cache?}
        B -->|Hit| C[Return]
        B -->|Miss| D[LLM]
        D --> E[Store]
        E --> C
    end

    classDef request fill:#6366F1,stroke:#7C90A0,color:#fff
    classDef cache fill:#F59E0B,stroke:#7C90A0,color:#fff
    classDef result fill:#10B981,stroke:#7C90A0,color:#fff

    class A request
    class B,E cache
    class C,D result

Enable Caching¶

Caching is enabled by default:

from testagent import TestConfig, set_config

set_config(TestConfig(cache_enabled=True))

Cache Location¶

Default: .testagent_cache/

set_config(TestConfig(cache_dir="my_cache/"))

CLI Commands¶

Clear Cache¶

testagent cache-clear

View Statistics¶

testagent cache-stats

Output:

Cache Statistics
  Entries: 42
  Size: 15360 bytes
  Location: .testagent_cache

TestAgentCache Class¶

Direct cache access:

from testagent import TestAgentCache, CacheKey

cache = TestAgentCache()

# Create cache key
key = CacheKey(
    output="test output",
    criteria="is correct",
    model="gpt-4o-mini"
)

# Get/set
result = cache.get(key)
cache.set(key, result)

# Statistics
stats = cache.stats()
print(f"Entries: {stats['entries']}")

# Clear
cache.clear()

CacheKey¶

Field	Type	Description
`output`	`str`	Tested output
`criteria`	`str`	Evaluation criteria
`expected`	`str`	Expected output
`model`	`str`	LLM model used

Best Practices¶

When to Cache

Repeated tests with same inputs
CI/CD pipelines
Development iterations

When Not to Cache

Testing non-deterministic outputs
When you need fresh evaluations