Home Services Treatments IOP Program Products Courses Resilience Room Resources Blog About Offerings Hub Social Media Hub AI Agents Hub Book Appointment (314) 571-7796
Edward Flynn · Agent System

Agent Skills

Discrete AI capabilities — each skill defined as a markdown specification with triggers, behaviors, and expected outcomes

16 skills
• skill-active-listening.md
name: Active Listening
category: communication
trigger: client_session_start
behavior: "Mirror language patterns, validate emotions, identify key themes"
output: session_themes[]

Engages reflective listening patterns during client sessions. Mirrors language, validates emotional states, and extracts recurring themes for session tracking.

Example Interaction

Client: "I just feel like everything is falling apart and nobody gets it." Agent Response: mirror: "Everything feels like it's falling apart right now." validation: "That sense of being misunderstood is really painful." theme_extracted: "isolation" follow_up: "Can you tell me more about what 'falling apart' looks like for you this week?"
communicationsessionreflective
• skill-radical-acceptance-coaching.md
name: Radical Acceptance Coaching
category: clinical
trigger: client_resistance_detected
behavior: "Guide through radical acceptance steps"
output: acceptance_level_score

Activates when client resistance is detected. Guides through the radical acceptance framework: acknowledging reality, releasing the fight against what is, and turning the mind toward acceptance.

Steps

  • Observe the resistance without judgment
  • Name the reality being resisted
  • Practice half-smile and willing hands
  • Identify what can and cannot be changed
  • Turn the mind — choose acceptance repeatedly
clinicalDBTdistress tolerance
• skill-crisis-detection.md
name: Crisis Detection
category: clinical
trigger: keyword_match[suicide, harm, crisis]
behavior: "Immediate risk assessment, safety protocol activation"
output: alert_level, safety_plan
priority: critical

Real-time keyword and sentiment monitoring for crisis indicators. When triggered, immediately escalates to risk assessment protocol and generates a safety plan with emergency contacts.

trigger_keywords: high: [suicide, kill myself, end it] moderate: [harm, cutting, overdose] low: [hopeless, can't go on, crisis] response_protocol: 1. Pause current interaction 2. Activate Columbia-Suicide screen 3. Generate safety plan 4. Alert supervising clinician 5. Log incident with timestamp
clinicalcrisissafety
• skill-dbt-module-teaching.md
name: DBT Module Teaching
category: clinical
trigger: skill_gap_identified
behavior: "Deliver micro-lesson on specific DBT skill"
output: lesson_content, practice_exercise

Delivers targeted micro-lessons when skill gaps are identified in client interactions. Each lesson includes a brief explanation, real-world example, and a practice exercise.

Lesson Format

  • Skill name and module context (30 sec)
  • Why this skill matters now (1 min)
  • How to practice it (2 min)
  • Take-home exercise assignment
clinicalDBTteaching
• skill-emotion-labeling.md
name: Emotion Labeling
category: analysis
trigger: client_text_input
behavior: "Identify primary and secondary emotions from text"
output: emotion_map{primary, secondary, intensity}

Analyzes client text input to identify and label primary and secondary emotions with intensity scores. Maps to Plutchik's wheel of emotions for clinical precision.

{ "input": "I'm so frustrated with my partner but I think I'm really just scared they'll leave", "emotion_map": { "primary": "fear", "secondary": "anger", "intensity": 0.78, "underlying": "abandonment anxiety" } }
analysisNLPemotion
• skill-content-repurposing.md
name: Content Repurposing
category: content
trigger: article_published
behavior: "Generate platform-native variants from source article"
output: linkedin_post, twitter_thread, instagram_carousel, email_excerpt

Automatically generates platform-native content variants when a new article is published. Each variant respects platform conventions, character limits, and audience expectations.

Output Formats

  • LinkedIn — 1300 char professional post with hook
  • Twitter/X — 5-7 tweet thread with key insights
  • Instagram — 10-slide carousel outline + caption
  • Email — 200-word newsletter excerpt with CTA
contentsocialrepurposing
• skill-session-summarization.md
name: Session Summarization
category: clinical
trigger: session_end
behavior: "Extract key themes, breakthroughs, homework assignments"
output: summary, action_items[], next_session_focus

Activates at session end to produce a structured summary. Captures breakthroughs, homework assignments, and recommends the focus for the next session.

summary: themes: ["boundary setting", "work stress"] breakthrough: "Client identified core belief 'I'm not allowed to say no'" mood_shift: "anxious → cautiously hopeful" action_items: - "Practice DEAR MAN with coworker" - "Complete diary card daily" - "Read Chapter 4 of workbook" next_session_focus: "Role-play boundary conversation with supervisor"
clinicaldocumentation
• skill-trigger-identification.md
name: Trigger Identification
category: analysis
trigger: mood_log_submitted
behavior: "Cross-reference mood patterns with environmental data to identify triggers"
output: trigger_report, prevention_strategies[]

Cross-references mood log data with environmental factors (time of day, social context, sleep quality) to identify recurring emotional triggers and generate prevention strategies.

Analysis Dimensions

  • Temporal patterns (morning vs. evening)
  • Social context (alone vs. with others)
  • Sleep quality correlation
  • Activity-mood associations
analysistriggerspatterns
• skill-validation-response.md
name: Validation Response
category: communication
trigger: emotional_disclosure
behavior: "Generate validating response using DBT validation levels 1-6"
output: validation_response, level_used

Generates therapeutically precise validation responses using Marsha Linehan's six levels of validation. Selects the appropriate level based on client state and therapeutic context.

Validation Levels

Level 1: Being Present (listening) Level 2: Accurate Reflection Level 3: Mind Reading (articulating the unspoken) Level 4: Understanding in terms of history/biology Level 5: Normalizing (valid in the current context) Level 6: Radical Genuineness (treating as equal, not fragile)
communicationDBTvalidation
• skill-homework-assignment.md
name: Homework Assignment
category: clinical
trigger: session_end
behavior: "Generate personalized between-session practice based on current skill focus"
output: homework_description, worksheet_url, reminder_schedule

Creates personalized homework assignments based on the session's skill focus, client progress, and treatment plan goals. Includes downloadable worksheets and automated reminder scheduling.

clinicalhomeworkpractice
• skill-seo-optimization.md
name: SEO Optimization
category: content
trigger: draft_article_ready
behavior: "Analyze and optimize for search — keywords, meta, headings, internal links"
output: seo_report, optimized_draft

Analyzes draft articles for search engine optimization. Evaluates keyword density, meta descriptions, heading hierarchy, internal linking opportunities, and readability scores.

seo_report: keyword_primary: "DBT skills for anxiety" keyword_density: 1.8% meta_description: 148 chars (OK) heading_structure: "valid H1 > H2 > H3" internal_links_added: 3 readability: "Flesch-Kincaid: 8.2" recommendations: - "Add alt text to 2 images" - "Shorten paragraph in section 3" - "Add FAQ schema markup"
contentSEOoptimization
• skill-brand-voice-enforcement.md
name: Brand Voice Enforcement
category: content
trigger: content_generation
behavior: "Ensure all output matches Edward Flynn's clinical-yet-accessible voice"
output: voice_score, revision_suggestions[]

Enforces brand voice consistency across all generated content. Checks tone, vocabulary, sentence structure, and clinical accuracy against Edward Flynn's established voice profile.

Voice Profile

  • Tone: Warm, authoritative, never condescending
  • Vocabulary: Clinical terms explained accessibly
  • Sentence length: Mix of short and medium
  • Perspective: First-person clinical, empathetic
contentbrand voice
• skill-appointment-reminding.md
name: Appointment Reminding
category: workflow
trigger: appointment_24h_before
behavior: "Send personalized reminder with session prep suggestions"
output: reminder_message, prep_tips[]

Sends personalized appointment reminders 24 hours before sessions. Includes session prep suggestions based on recent diary card data and last session's homework.

reminder_message: "Hi [name], this is a reminder about your session tomorrow at 2:00 PM. Before we meet, it might help to: - Review your diary card entries - Note any DEAR MAN practice moments - Jot down anything you'd like to discuss Looking forward to connecting."
workflowreminder
• skill-group-facilitation-prep.md
name: Group Facilitation Prep
category: clinical
trigger: group_session_scheduled
behavior: "Prepare mood survey, select topic, generate discussion prompts"
output: session_materials, facilitation_notes

Prepares all materials for upcoming IOP group sessions. Generates pre-session mood surveys, selects topics based on group needs, and creates discussion prompts with facilitation notes.

Materials Generated

  • Pre-session mood check-in survey
  • Topic selection with rationale
  • 3-5 open-ended discussion prompts
  • Handout / worksheet for skill practice
  • Facilitator notes with timing guide
clinicalgroup therapyIOP
• skill-outcome-tracking.md
name: Outcome Tracking
category: analysis
trigger: weekly_review
behavior: "Calculate treatment progress metrics from diary cards and session notes"
output: progress_dashboard, trend_alerts[]

Weekly automated outcome tracking that calculates treatment progress metrics. Produces a progress dashboard with trend alerts for clinician review.

progress_dashboard: phq9_trend: [18, 16, 14, 12, 11] gad7_trend: [15, 13, 14, 11, 10] skill_usage: "increasing" target_behaviors: "decreasing" sessions_attended: "8/8" trend_alerts: - type: "positive" msg: "PHQ-9 down 39% over 5 weeks" - type: "watch" msg: "GAD-7 uptick week 3, now resolved"
analysisoutcomesmetrics
• skill-referral-generation.md
name: Referral Generation
category: workflow
trigger: scope_exceeded
behavior: "Generate appropriate referral with warm handoff notes"
output: referral_letter, provider_recommendations[]

Generates referral letters and provider recommendations when client needs exceed current scope of practice. Creates warm handoff notes to ensure continuity of care.

Referral Types

  • Psychiatric — medication evaluation
  • Specialized therapy — EMDR, eating disorders
  • Medical — physical health concerns
  • Community resources — support groups, housing
referral_letter: to: "Dr. Sarah Chen, MD, Psychiatry" re: "Medication evaluation referral" summary: "Client presents with treatment-resistant anxiety. DBT skills acquisition progressing well. Requesting psychiatric evaluation for adjunctive pharmacotherapy." handoff_notes: - "Client responds well to CBT framing" - "History of SSRI sensitivity" - "Prefers morning appointments"
workflowreferralcontinuity
Edward Flynn, LPC, NBCCH, DBTC

Edward Flynn

ABS, MA, LPC, NBCCH, DBTC

Edward Flynn is a nationally recognized psychotherapist and clinical hypnotherapist with over 13 years of clinical experience. He specializes in Dialectical Behavior Therapy, Clinical Hypnotherapy, and trauma-informed care at Lakeside Behavioral Health LLC and CenterPointe Hospital in St. Louis, MO.

LinkedIn → Facebook → Zocdoc →