Thinking Tools¶

The ThinkingTools toolkit provides a text-first thinking/journaling chain for agents:

• Build a step-by-step chain
• Add meta-cognitive reflections
• Store intermediate data in a scratchpad
• Synthesize the chain into a compact output

All public functions return strings (human-readable, markdown-ish).

🤖 AI Agent Setup (Agno)¶

from agno.agent import Agent
from enhancedtoolkits import ThinkingTools

agent = Agent(
    name="Planner",
    model="gpt-4",
    tools=[ThinkingTools(max_chain_length=10, confidence_threshold=0.7)],
)

The agent parameter is used as the session state container.

🔧 Valid values¶

• thinking_type: analysis | synthesis | evaluation | planning | creative | reflection
• synthesis_type: conclusion | summary | insights | next_steps
• scratchpad operation: set | get | list | clear

🧠 Available Functions¶

build_step_by_step_reasoning_chain(agent, problem, thinking_type='analysis', context=None, evidence=None, confidence=0.5)¶

Starts a chain (first call) or appends a step (subsequent calls).

add_meta_cognitive_reflection(agent, reflection, step_id=None)¶

Adds a reflection entry for the current chain.

manage_working_memory_scratchpad(agent, key, value=None, operation='set')¶

Scratchpad operations: - set: set key=value - get: retrieve one key - list: list all keys - clear: clear one key or all keys (key='all')

assess_reasoning_chain_quality_and_suggest_improvements(agent)¶

Computes a compact quality score and suggests improvements.

synthesize_reasoning_chain_into_output(agent, synthesis_type='conclusion')¶

Returns a short synthesis and closes the current chain.

retrieve_current_thinking_chain_state(agent)¶

Returns a compact state summary.

reset_current_thinking_chain(agent)¶

Clears the current chain (does not delete history summaries).

✅ Example Workflow¶

from enhancedtoolkits import ThinkingTools

thinking = ThinkingTools()

# 1) Start a chain
thinking.build_step_by_step_reasoning_chain(
    agent=agent,
    problem="Plan a 2-week study schedule for system design interviews",
    thinking_type="planning",
    context="I can study 90 minutes per weekday and 3h on weekends",
    confidence=0.6,
)

# 2) Add another step
thinking.build_step_by_step_reasoning_chain(
    agent=agent,
    problem="Week 1: focus on scalability fundamentals + 2 mock designs",
    thinking_type="planning",
    confidence=0.7,
)

# 3) Add a reflection
thinking.add_meta_cognitive_reflection(
    agent=agent,
    reflection="Risk: underestimating review time; schedule buffer slots.",
)

# 4) Synthesize into next steps
thinking.synthesize_reasoning_chain_into_output(
    agent=agent,
    synthesis_type="next_steps",
)

API Reference¶

• docs/api/thinking.md