{% extends "base.html" %} {% block title %}Cognitive State — Agent Wasp{% endblock %} {% block page_css %} {% endblock %} {% block content %}
The agent's living memory and belief systems
{{ title }}
{% if f.frequency or f.risk %}{% if f.frequency %}{{ f.frequency }}{% endif %} {% if f.frequency and f.risk %} · {% endif %} {% if f.risk %}{{ f.risk }} risk{% endif %}
{% endif %} {% if f.solution %}→ {{ f.solution }}
{% endif %}No failures recorded
{% endfor %}{{ sr.success }}✓ {{ sr.failure }}✗ · {{ sr.total }} total
{{ epistemic.get('total_interactions', 0) }} total interactions {% if epistemic.get('last_updated') %} · updated {{ epistemic['last_updated'][:10] }}{% endif %}
No epistemic data yet
{% endfor %}Last check: {{ gen_at[:19] | replace("T", " ") }}
{% endif %} {% if integrity_report.get('corrections') %}✓ {{ c }}
{% endfor %}{{ f.detail }}
{% if f.type == 'epistemic_drift' %}No integrity issues found
All self-model claims are consistent with observed reality.
No integrity report available yet.
The Self-Integrity Monitor runs every 6 hours.
Pending code changes proposed by the agent
{{ p.change }}
{% endif %} {% if p.diff_preview %}{{ p.diff_preview }}{% if p.diff_preview|length >= 300 %}…{% endif %}
{% endif %}
No pending proposals
The agent stores code change proposals here before they're reviewed.
Learned from user corrections · injected into every prompt
{{ rule.description }}
No behavioral rules yet
Created when the agent detects and processes user corrections.
Few-shot examples driving behavior · sorted by usage
{{ ex.user_input }}
{% if ex.skill_calls %}→ {{ ex.skill_calls }}
{% endif %}No learning examples yet
Stored automatically from user feedback interactions.