AI Skill Report Card

Reframing Inner Critic

A-82·Nov 10, 2026·Source: Web
14 / 15

Capture the exact critical phrase, then reframe it immediately:

Critical phrase: "Я опять всё забыл, я бестолковый" Reframed: "Моя память перегружена, это нормально. Давай я запишу это в заметки и поставлю напоминание."

Pattern: acknowledge the fact → normalize it → offer a practical next step.

Recommendation▾
Include at least one English-language example alongside the Russian ones to make the skill more broadly applicable/accessible
14 / 15

Progress:

  • Step 1: Capture the negative inner dialogue phrase verbatim (don't paraphrase yet)
  • Step 2: Identify the harsh elements (labeling, overgeneralization, catastrophizing)
  • Step 3: Rewrite as a wise, benevolent, objective mentor would say it
  • Step 4: Check the reframe has 3 parts: fact acknowledgment, normalization, actionable step
  • Step 5: Offer the phrase back and ask the person to say/write it themselves
  • Step 6: Suggest journaling the before/after pair for future reference
  • Step 7: Encourage independent use of this pattern without AI next time

Step 1: Capture the phrase

Ask the person to state exactly what they said to themselves, including tone. Don't soften it yet — the raw version is needed to identify distortion patterns.

Step 2: Identify the distortion

Common patterns to name:

  • Labeling: "я бестолковый", "я неудачник" (identity-level judgment from a single event)
  • Overgeneralization: "я ВСЕГДА всё порчу", "опять" implying a permanent pattern
  • Catastrophizing: treating a minor lapse as evidence of deep inadequacy
  • All-or-nothing: no room for "normal human variation"

Step 3: Reframe using mentor voice

The mentor voice is:

  • Factual, not dismissive — doesn't deny the event happened
  • Compassionate, not indulgent — no "it's totally fine, don't worry about it ever"
  • Forward-looking — ends in something actionable, not just comfort

Template:

[Nonjudgmental restatement of the fact] + [normalizing context] + [concrete next step]

Step 4: Verify the three parts are present

A good reframe is NOT just "be kind to yourself" — it must retain respect for reality and agency. Reject reframes that are purely soothing with no action, and reject ones that still carry judgment disguised as "honesty."

Step 5-7: Hand it back

The goal is NOT dependency on the AI. Explicitly tell the user: "Next time, try generating this reframe yourself using the same three-part structure before reaching out."

Recommendation▾
Add an example of a 'bad' reframe (e.g., toxic positivity version) contrasted with the good one to reinforce the pitfall section concretely
15 / 20

Example 1: Input: "Я опять облажался на созвоне, вечно я всё порчу" Output: "На этом созвоне я не донёс мысль так, как хотел — это случается. Это не 'вечно', это конкретный момент. В следующий раз я могу набросать тезисы заранее."

Example 2: Input: "Я ленивая, весь день ничего не сделала" Output: "Сегодня был день с низкой энергией, и я не выполнила план. Тело и разум иногда требуют паузы — это не лень, это сигнал. Завтра я начну с одной маленькой задачи."

Example 3: Input: "Я тупой, опять не понял с первого раза" Output: "Мне потребовалось больше одной попытки, чтобы разобраться — это нормальная часть обучения, а не показатель ума. Могу перечитать ещё раз или попросить объяснить иначе."

Recommendation▾
Clarify how to handle cases where the person resists reframing or insists the harsh self-talk is 'accurate'
  • Keep the reframe short enough to actually say to yourself in the moment — long paragraphs won't be used in real time
  • Always preserve the factual core of the original complaint — the goal is reframing, not denial
  • End with agency (a small next step), not just comfort — this is what separates mentor tone from empty positivity
  • Encourage the person to build their own internal library of reframed phrases for recurring patterns (memory, deadlines, social anxiety, etc.)
  • Suggest a simple journal format: date | trigger | critical phrase | reframed phrase
  • Tie reframes to CBT distortion names so the person learns to self-diagnose over time (labeling, overgeneralization, catastrophizing, mind-reading, should-statements)
  • Toxic positivity: "Everything is fine, don't be so hard on yourself!" — invalidates the real issue and feels hollow
  • Losing the action step: a reframe that only comforts without suggesting what to do next doesn't build real self-support skills
  • Making it too long: an elaborate therapeutic paragraph won't be repeated in the heat of the moment — brevity matters
  • Skipping Step 1: reframing before clearly capturing the original phrase leads to vague, generic advice instead of targeted mentor language
  • Creating dependency: always nudge the person toward generating their own reframes — the skill is meant to be internalized, not outsourced permanently
  • Ignoring recurring patterns: if the same critical phrase appears repeatedly, flag it and suggest exploring the deeper belief behind it (e.g., recurring "I'm stupid" may point to a core belief worth separate reflection)
0
Grade A-AI Skill Framework
Scorecard
Criteria Breakdown
Quick Start
14/15
Workflow
14/15
Examples
15/20
Completeness
17/20
Format
15/15
Conciseness
13/15