Diagnose and adjust agent runtime parameters with strict validation.
npx skills add https://github.com/gongzeq/VAPT3 --skill my-gongzeq
Please help me install this Agent Skill. Skill: my Source: https://github.com/gongzeq/VAPT3/tree/main/secbot/skills/my Command: npx skills add https://github.com/gongzeq/VAPT3 --skill my-gongzeq
This skill helps diagnose why the agent won't perform as expected by inspecting and adjusting its runtime state, including the model, iterations, context window, token usage, and web configuration, and helps remember user preferences across turns.
Check the agent's current model and context window, then adjust max_iterations if the task is likely to require more steps.
High-intent search queries and answers about installing and using this skill.
Diagnose agent misbehavior by inspecting the runtime state to review current model, iterations, context window, and token usage. Checking these parameters helps identify misconfigurations causing unexpected agent performance.
Adjust max iterations and context window tokens via runtime state inspection. Enforce in-memory modifications with strict validation, ensuring max_iterations stays within 1–100 and context_window_tokens within 4096–1,000,000.
Agent runtime state includes the active model, iterations, context window, token usage, and web configuration. Inspect it when the agent won't perform as expected or when tuning resource limits to improve reliability.
Modifications are limited to in-memory changes only, requiring explicit safety checks before applying impactful settings. Protected parameters enforce strict validation: max_iterations must be 1–100, context_window_tokens 4096–1,000,000, and model must be non-empty.
Recall preferences across turns to maintain consistent agent behavior by storing and retrieving user settings within the runtime state. This ensures the agent applies previously established configurations in subsequent interactions.
trotsky1997Insert breakpoints and inspect runtime state in Python code.
panaversityValidate a Skill Unit against nine quality criteria and generate a structured report.
zylos-aiMonitor context window usage, token consumption, and session costs from statusLine JSON.
TeamDay-AIDeploy and configure AWS Bedrock AgentCore across Gateway, Runtime, Memory, and Identity services.
