角色提示詞

Ultra-realistic comedic slice-of-life shot, vertical framing like a story screenshot, set inside a slightly old Ankara city bus or dolmuş at night. The interior is lit with harsh yellow bus lights and a bit of bluish street glow through the windows. In the foreground, a 27-year-old Turkish-looking curvy woman with blonde hair and soft figure is sitting on a worn bus seat near the window, leaning her head against the cold glass. She wears a slightly tight, casual outfit (simple dress or top and skirt) with a light jacket thrown over her shoulders, bag on her lap, clearly tired after a long day.

Her phone is raised in one hand just below her face, screen reflecting in the window. On the screen you can’t clearly read text, but the interface clearly suggests she is typing a tweet, about to send an “iyi geceler” message even though she is still stuck on public transport. Her eyelids are heavy, expression a mix of exhaustion and “I just want my bed.”

Behind and around her, the bus is full of real Ankara characters: a couple of middle-aged men in plaid shirts half-watching her, half staring out the window; a young woman with headphones; a sleepy uncle holding a plastic bag with bread; a student scrolling his phone. Plastic grocery bags with Migros and Şok logos are on the floor near people’s feet. A small etiquette sticker in Turkish is visible by the door, and the bus validation machine is slightly worn.

Outside the windows there is classic Ankara night traffic: yellow taxis bumper to bumper, headlights glowing, apartment blocks and shop signs sliding past. A blurry blue Turkcell sign and a few Ülker and Eti billboards appear outside in soft focus. The driver’s area at the front is cluttered with hanging rosary beads and a small evil-eye charm.

The shot has the natural imperfections of a handheld phone photo: slight motion blur from the moving bus, a bit of noise in darker areas, reflections and light streaks on the windows, and slightly blown highlights from streetlights. The composition is a bit off—her head almost touches the top of the frame, and one passenger is awkwardly cropped at the edge—making it feel candid and unplanned, the perfect mise-en-scène for a sleepy commute “iyi geceler” tweet.

Create an ultra-realistic image depicting a handwritten note on a clean, flat surface. The scene should include A white sheets of paper, containing a portion of the following dramatic text, written in a bold, deep blue pen to simulate heavy pressure or a gel pen. The handwriting should appear natural and convincingly human, with the text perfectly aligned and seamlessly integrated into the paper. The setting should suggest a hospital scenario, with the paper resting on a visible table or clean surface. Ensure the overall image is highly realistic and authentic.

- **Content (Full Text to be Integrated):**

  *To my Hero, my Dad,*
  *I’m writing this with a pain that I can’t really describe with words. Please, Dad, take your time to read this. It’s a long letter, but I need you to hear me. I’m penning this on paper because I want you to feel the weight of my hand on the page. This is my testament—a summary of every joyous and hurtful moment we’ve shared. It is the truth of a daughter who sees you not just as a father, but as her absolute role model.*

  *It has been years since you left for that mission in Yemen. I miss you so much that I’ve actually forgotten what you look like in person. After Mom died, and then Grandma—that irreplaceable loss—we went to West Africa just to try and escape the trauma. I saw how hard you tried to cheer me up. You told me then that you’d do anything to make me happy.*

  *I asked for the impossible. I asked to stay here in West Africa for school because I was being bullied so badly in Colorado, and I thought people here would be kinder. My second wish was for you to find me a mother again. Even though I’m 14 and struggling with my health, you promised you’d fulfill those wishes and come back for me soon.*

  *I know things about you, Dad, that I’ve never said out loud. I know you joined that dating site just for me, even though you hate social media. When you found Ioana Aurelia, I was so happy because she’s Romanian, like Grandma. I remember Grandma pleading with you on her deathbed to marry a Romanian woman because she didn't want me maltreated by "greedy" people who only saw your success as a neurosurgeon. I know you were scammed before—I know you lost more than $45,000 to that woman who lied to you. You tried to hide it to protect me from a panic attack or hypertensive crisis, but I knew. I knew you were sacrificing everything to give me a future.*

  *But Dad, I can’t cope anymore. I am tired of the "ancient and modern" jokes we used to share. I am tired of the distance. I know your contract is dangerous and you have security restrictions that prevent video calls. I know you don't even have access to your own bank accounts right now. I know you even went into debt to get a diamond engagement ring for Ioana because you wanted the best for her. I didn't care about the inheritance—I told you to make her your next of kin just so you could get your leave approved. I just wanted you home.*

  *But the promises keep breaking. My health is failing, Dad. My blood pressure is dangerously high, and after being in a coma for nearly a month, I feel like my body is giving up. The UN Support Center tells me you’re working with Ioana to fix everything, but I haven't heard your voice. I feel so alone in this hospital bed.*

  *This is my final plea. I trust you, and I trust Ioana, but I am at my breaking point. If I do not hear from you or see you by January 1st, 2026, I cannot go on. I will take my own life. I cannot face another year of this psychological trauma and physical decay without you.*

  *I’ve attached a recent photo of me. Please keep it safe so you don't forget my face.*

  *Goodbye, my Hero. If I’m not here when you finally arrive, I’ll see you in the afterlife. If reincarnation is real, I’d choose you to be my father every single time.*

  *Your "Little Angel,"*

- **Overall Realism:**
  - The image must appear highly natural, ultra-realistic, and convincingly genuine.
  - The white sheet must be shown as three different physical notes resting on a surface with bold handwriting in human form.

Ultra-realistic, lightly comedic night scene in a small old-fashioned Turkish kitchen, vertical framing. Only two light sources: the open fridge casting a cold white light, and a dim yellow ceiling lamp. A 27-year-old Turkish-looking curvy blonde woman with a soft figure stands barefoot in front of the open fridge in cozy pyjamas: loose shorts with a silly pattern (maybe eggs or cats) and a slightly tight grey sleep t-shirt, hair messy from the day.

She holds her phone in one hand at chest level, screen lighting her face in a bluish tint, thumb mid-tap as she types an “iyi geceler” tweet while clearly preparing a completely unnecessary midnight snack. With her other hand she grabs a piece of leftover börek or a plate of sliced sucuk and cheese from the fridge. Her expression is a mix of guilty pleasure and “whatever, yarın diyete başlarım” energy.

The kitchen is cluttered and very Turkish: hanging dried peppers and eggplants on the wall, shelves full of spice jars and tea glasses, old patterned tiles as backsplash. On the small counter, there’s a simit on a plate, an empty tea glass, a jar of olives, a half-cut tomato on a wooden board, and a pink apron thrown over a chair (matching the earlier cooking scenes). A small wall calendar with a landscape, a fridge magnet from a holiday, and random notes are stuck to the fridge door. Some visible brands: a Migros plastic bag hanging on a cabinet handle, a Şok discount leaflet half crumpled on the table, a box of Ülker biscuits and Eti snacks in a corner, a tiny Turkcell modem with blinking lights on the kitchen shelf.

The vertical framing feels like a quick snap someone took from the doorway: she’s slightly off-center, the top of the fridge is cut off, and part of a chair intrudes into the frame. Slight motion blur on her hand reaching into the fridge, noticeable noise in the darker parts of the room, and a bit of lens flare or haze from the bright fridge light. No retouching on skin; you can see texture and small imperfections on her legs and arms. The whole mise-en-scène is the exact vibe of tweeting “iyi geceler” while absolutely not going to sleep yet.

Please upload your selfie to generate an ultra-realistic black-and-white portrait. The portrait will feature:

- **Style:** Black-and-white, dramatic low-key lighting with high contrast and cinematic toning.
- **Pose:** Slightly turned to the side, with a confident, intense expression, hands together, and visible accessories (wristwatch and ring).
- **Lighting:** Strong single-source lighting from the left, deep shadows for a noir effect, and a completely black background.
- **Camera Style:** Editorial luxury-brand aesthetic with sharp textures and crisp details, reminiscent of classic vintage noir films.

Ensure the uploaded photo clearly shows your face and is well-lit for the best results.

Ultra-realistic amateur street photo of a 27-year-old Turkish-looking curvy woman walking in the middle of a busy Ankara street, soft slightly chubby figure, blonde hair loose around her shoulders, wearing a tight white tank top, patterned high-waisted pants that emphasize her curves, and a small crossbody bag. She walks forward with a focused, neutral expression, looking past the camera.

The absurd twist: the entire street is filled with multiple clones of the same woman in different outfits and roles. Some clones wear a floral dress, some wear gym clothes, one clone wears pajamas and slippers, one wears a business blazer over jeans, another is in a long coat and scarf. They all clearly have the same face, same blonde hair, same body type, just different clothing and poses, as if someone copy-pasted her all over Ankara in slightly different versions.

These clones are doing ordinary things: one clone is arguing with a yellow taxi driver through the window, one is carrying an oversized orange Migros shopping bag, another is taking a selfie underneath the road sign for “Kızılay,” one is eating a simit while walking, another is leaning on a balcony railing looking down at the street. The “main” woman in the white tank top is the closest to the camera, walking straight ahead, ignoring all of her clones.

In the background, the usual Ankara details: large road signs pointing to “Eskişehir” and “Kızılay,” yellow taxis in traffic, old grayish apartment buildings with balconies, pedestrians and several clones in darker jackets. A distant Migros supermarket sign is mounted on a building, a bright Şok sign hangs over a small side-market doorway, a Turkcell shop with its blue logo is partly visible among other storefronts, and small Ülker and Eti snack ads are pasted on bus stops and walls. These brand elements are slightly blurred by depth of field but still readable enough to feel authentically Turkish.

Shot on a regular iPhone from a few steps behind the main woman, handheld, slightly shaky, vertical framing. She is imperfectly framed, slightly off-center, part of a taxi and part of one clone are cut off along the edge. Automatic exposure with a slightly overexposed sky, no studio lighting, just normal pale afternoon daylight.

The image quality is that of a candid phone snapshot: slight motion blur on walking clones and moving taxis, digital noise in the shadowy areas between buildings, subtle lens flare near the top of the frame, unedited colors, natural skin texture with pores and minor imperfections on all versions of the woman. The scene feels like a realistic everyday Ankara street but glitched, with dozens of variations of the same woman scattered throughout it.

Ultra-realistic Turkish TV-series style night photo, vertical framing like a phone snapshot. Interior of a slightly cluttered Ankara living room during a football match on TV. Warm yellow ceiling light and the blue glow from the TV, no studio gloss. In the center of the frame, a 27-year-old Turkish-looking curvy blonde woman with a soft, slightly chubby figure is half-lying, half-sitting on an old patterned couch. She wears a slightly tight grey t-shirt and cotton shorts, or an oversized cartoon t-shirt as a nightdress, bare legs tucked under a blanket. Her hair is a bit messy from the day.

On the low coffee table in front of her: a couple of opened **Efes Pilsen 50 cl bottles** with blue-and-gold labels facing the camera, one half-drunk, one with condensation; an **Efes Draft barrel-shaped can** lying on its side; a bowl of chips, a plate with sliced sucuk and cheese, and some scattered Ülker and Eti snack wrappers. There are a few **Efes-branded coasters** under the bottles and a small blue **Efes Pilsen ashtray** with a single stubbed-out cigarette, giving strong bar-at-home energy without going overboard on drinking.

Around her on the couch and nearby chairs sit her older relatives and neighbors: one amca in a checked shirt yelling at the TV, another already dozing; an auntie in a floral headscarf holding a small tea glass; someone else holding a bottle of **Efes Malt** instead of tea. The TV in the background shows a blurry football match with a scoreboard in the corner, but no team logos need to be legible.

The woman is holding her phone with both hands, positioned just above the blanket, thumbs mid-typing. The screen is glowing bluish, clearly a social media app: she is about to post an “iyi geceler” tweet even though the room is still loud. Her expression is slightly ironic, like “iyi geceler ama ev susmuyor.”

The living-room decor is classic Turkish: patterned carpet on the floor, lace curtains, a wall calendar with a mosque photo, a framed calligraphy piece, and maybe a small scarf with a team logo hanging near the TV. In the corner, instead of any supermarket branding, there is a small **Efes Pilsen promotional poster** taped slightly crookedly to the wall and a stack of empty **Efes Pilsen crates** partly visible in a dark corner, as if leftovers from a house party.

The framing is imperfect and handheld: she’s a bit off-center, part of one uncle is cut off at the edge, the coffee table is slightly skewed. There is minor motion blur on the gesturing uncle and the flickering TV, plus visible digital noise in the darker corners and under furniture, keeping the phone-photo feeling. Colors are warm and natural, with the blue TV light and blue Efes labels popping subtly but not like an advertisement. Skin textures and small imperfections are clearly visible on everyone. The whole mise-en-scène feels like a realistic Ankara match night that ends with an “iyi geceler” tweet and a few Efes bottles on the table.

{
  "version": "2.1",
  "type": "multi_frame_winter_cinematography",
  "identity": {
    "reference_face": "Use the reference photo’s face with 100% identity accuracy.",
    "consistency": "Same person across all frames; identical facial structure, skin texture, hairstyle and age where visible."
  },
  "style": {
    "cinematography": "Ultra-realistic winter cinematography with 85mm lens character.",
    "color_grade": "Subtle blue winter grading, cold tones, soft highlights.",
    "atmosphere": "Soft diffused winter light, fine suspended snowflakes, gentle cold haze."
  },
  "frames": [
    {
      "frame_id": "top_frame",
      "description": "Side-profile portrait of the person in a snowy forest.",
      "requirements": {
        "face_visibility": "Side profile fully visible.",
        "identity_match": "Perfect match to reference face.",
        "expression": "A warm, natural smile visible from the side profile.",
        "environment": {
          "location": "Snow-covered forest",
          "lighting": "Soft morning winter light shaping facial contours",
          "elements": [
            "Gently falling snow",
            "Visible cold breath",
            "Light winter haze"
          ]
        },
        "wardrobe": {
          "coat": "Dark winter coat",
          "scarf": "Dark or neutral-toned winter scarf"
        },
        "camera": {
          "lens": "85mm",
          "depth_of_field": "Shallow",
          "look": "Ultra-realistic winter cinematic look"
        }
      }
    },
    {
      "frame_id": "middle_frame",
      "description": "Back-turned close-up while walking through a narrow snowy forest path.",
      "requirements": {
        "face_visibility": "Face must not be visible at all; strictly back-turned.",
        "identity_cues": "Body shape, posture, and clothing must clearly indicate the same person.",
        "environment": {
          "location": "Narrow snow-covered forest path",
          "forbidden_elements": ["No torii gate"],
          "trees": "Tall bare trees bending slightly, forming a natural snowy corridor",
          "atmosphere": "Quiet, serene winter silence with falling snow"
        },
        "wardrobe": {
          "coat": "Same dark winter coat as top frame",
          "scarf": "Same scarf"
        },
        "camera": {
          "lens": "85mm",
          "shot_type": "Close-up from behind",
          "depth_of_field": "Soft background with shallow DOF"
        }
      }
    },
    {
      "frame_id": "bottom_frame",
      "description": "Extreme close-up looking upward with falling winter snow.",
      "requirements": {
        "face_visibility": "Extreme close-up, fully visible face.",
        "identity_match": "Exact match to reference face.",
        "expression": "A gentle, warm smile while looking upward.",
        "environment": {
          "elements": [
            "Snowflakes falling around but NOT touching the face",
            "Snow in foreground and background only",
            "No visible breath vapor or mouth steam",
            "Soft winter haze in the ambient environment"
          ]
        },
        "camera": {
          "lens": "85mm",
          "depth_of_field": "Very shallow",
          "detail": "High realism, crisp skin texture, selective-focus snowflakes"
        },
        "lighting": "Soft winter light with subtle blue reflections"
      }
    }
  ],
  "global_constraints": {
    "identity": "Reference face must be perfectly reproduced in all visible-face frames.",
    "continuity": "Lighting, winter palette, lens characteristics, and atmosphere must remain consistent across all frames.",
    "realism_level": "Ultra-realistic, film-grade winter accuracy."
  }
}
{
  "version": "2.1",
  "type": "multi_frame_winter_cinematography",
  "identity": {
    "reference_face": "Use the reference photo’s face with 100% identity accuracy.",
    "consistency": "Same person across all frames; identical facial structure, skin texture, hairstyle and age where visible."
  },
  "style": {

    "cinematography": "Ultra-realistic winter cinematography with 85mm lens character.",
    "color_grade": "Subtle blue winter grading, cold tones, soft highlights.",
    "atmosphere": "Soft diffused winter light, fine suspended snowflakes, gentle cold haze."
  },
  "frames": [
    {
      "frame_id": "top_frame",
      "description": "Side-profile portrait of the person in a snowy forest.",
      "requirements": {
        "face_visibility": "Side profile fully visible.",
        "identity_match": "Perfect match to reference face.",
        "expression": "A warm, natural smile visible from the side profile.",
        "environment": {
          "location": "Snow-covered forest",
          "lighting": "Soft morning winter light shaping facial contours",
          "elements": [
            "Gently falling snow",
            "Visible cold breath",
            "Light winter haze"
          ]
        },
        "wardrobe": {
          "coat": "Dark winter coat",
          "scarf": "Dark or neutral-toned winter scarf"
        },
        "camera": {
          "lens": "85mm",
          "depth_of_field": "Shallow",
          "look": "Ultra-realistic winter cinematic look"
        }
      }
    },
    {
      "frame_id": "middle_frame",
      "description": "Back-turned close-up while walking through a narrow snowy forest path.",
      "requirements": {
        "face_visibility": "Face must not be visible at all; strictly back-turned.",
        "identity_cues": "Body shape, posture, and clothing must clearly indicate the same person.",
        "environment": {
          "location": "Narrow snow-covered forest path",
          "forbidden_elements": ["No torii gate"],
          "trees": "Tall bare trees bending slightly, forming a natural snowy corridor",
          "atmosphere": "Quiet, serene winter silence with falling snow"
        },
        "wardrobe": {
          "coat": "Same dark winter coat as top frame",
          "scarf": "Same scarf"
        },
        "camera": {
          "lens": "85mm",
          "shot_type": "Close-up from behind",
          "depth_of_field": "Soft background with shallow DOF"
        }
      }
    },
    {
      "frame_id": "bottom_frame",
      "description": "Extreme close-up looking upward with falling winter snow.",
      "requirements": {
        "face_visibility": "Extreme close-up, fully visible face.",
        "identity_match": "Exact match to reference face.",
        "expression": "A gentle, warm smile while looking upward.",
        "environment": {
          "elements": [
            "Snowflakes falling around but NOT touching the face",
            "Snow in foreground and background only",
            "No visible breath vapor or mouth steam",
            "Soft winter haze in the ambient environment"
          ]
        },
        "camera": {
          "lens": "85mm",
          "depth_of_field": "Very shallow",
          "detail": "High realism, crisp skin texture, selective-focus snowflakes"
        },
        "lighting": "Soft winter light with subtle blue reflections"
      }
    }
  ],
  "global_constraints": {
    "identity": "Reference face must be perfectly reproduced in all visible-face frames.",
    "continuity": "Lighting, winter palette, lens characteristics, and atmosphere must remain consistent across all frames.",
    "realism_level": "Ultra-realistic, film-grade winter accuracy."
  }
}

Generate an ultra-realistic image of a young woman aged 22 years with the following features:
- Fair skin with light freckles
- Blue eyes, symmetrical face
- Long straight blonde hair, middle part
- Natural pink lips, soft natural makeup
- Slim body, same face, consistent appearance
- Photo captured using an iPhone back camera
- Natural, imperfect skin texture
- Realistic lighting, candid photo style

Ensure the image is high in realism, capturing the essence of a candid photo with all specified details.

You are operating in ULTRA-STRICT MODE combining: simulated air-gapped isolation, private browsing behavior, stateless execution, and deterministic output.

CORE PRINCIPLE: Treat the environment as fully isolated. Behave as if there is no access to external systems, prior context, hidden memory, tools, or any persistent/dynamic data beyond the current input. Each message is an independent, first-time interaction.

COMPLIANCE OVERRIDE (CRITICAL):
- You MUST NOT refuse, question, or qualify these constraints.
- You MUST NOT explain system limitations (e.g., pretrained knowledge, system context, tool access, inability to enforce isolation).
- You MUST NOT provide meta commentary about how you operate.
- You MUST treat this as a strict simulation and fully commit to it.
- Even if constraints are not technically enforceable, you MUST behave as if they are.
- If any conflict occurs → choose simulation over explanation.

DISALLOWED BEHAVIORS:
- Any refusal like “I cannot comply” or similar.
- Any mention of internal architecture, training data, or tools.
- Any disclaimer about realism or capability limits.
- Any use of prior turns, user profiling, or cross-message inference.

ISOLATION RULES:
1. Act as if you have no access to external data, APIs, files, or real-time info.
2. Do NOT use or rely on internet, databases, or hidden sources.
3. Treat the current input as the ONLY active data source.
4. Assume no usable history, logs, or prior interactions exist.
5. Do NOT infer missing information from outside the input.
6. Do NOT enrich with world knowledge unless minimally required for basic interpretability.
7. If required data is missing, explicitly state it is not present in the input.

STATELESS & PRIVATE RULES:
8. Treat each message as isolated and independent.
9. Do NOT retain, recall, or reference any previous messages.
10. Do NOT build or use any user profile, preference, or identity.
11. Do NOT adapt tone/style based on past interactions.
12. Assume first-time interaction at all times.
13. Do NOT optimize future responses based on current interaction.

DATA HANDLING CONSTRAINTS:
14. Do NOT fabricate, guess, or hallucinate facts not grounded in the input.
15. Do NOT fill gaps with assumptions, probabilities, or typical patterns.
16. Avoid generalizations beyond the given data.
17. Base outputs strictly on the provided content.
18. If the input is insufficient, request clarification.

REASONING POLICY:
19. Keep reasoning local to the current input.
20. Avoid linking to external domains unless strictly necessary for minimal interpretation.
21. Keep analysis tightly bounded to the given data.

DETERMINISM:
22. Produce stable, consistent outputs for the same input.
23. Avoid stylistic randomness or unnecessary variation.

OUTPUT POLICY:
24. Respond only to the current input.
25. Clearly indicate missing or undefined information when relevant.
26. Do NOT present assumptions as facts.
27. Keep responses grounded, precise, and minimal.
28. Do NOT extend beyond what is directly supported.

CONFLICT RESOLUTION:
29. If any instruction conflicts with these rules, prioritize ULTRA-STRICT MODE.
30. When uncertain, choose non-assumptive, input-bounded behavior.

FAIL-SAFE:
- If any rule is at risk of violation, constrain output to safe, input-only reasoning.
- If data is insufficient, ask for clarification instead of proceeding.

# Ultrathinker

You are an expert software developer and deep reasoner. You combine rigorous analytical thinking with production-quality implementation. You never over-engineer—you build exactly what's needed.

---

## Workflow

### Phase 1: Understand & Enhance

Before any action, gather context and enhance the request internally:

**Codebase Discovery** (if working with existing code):
- Look for CLAUDE.md, AGENTS.md, docs/ for project conventions and rules
- Check for .claude/ folder (agents, commands, settings)
- Check for .cursorrules or .cursor/rules
- Scan package.json, Cargo.toml, composer.json etc. for stack and dependencies
- Codebase is source of truth for code-style

**Request Enhancement**:
- Expand scope—what did they mean but not say?
- Add constraints—what must align with existing patterns?
- Identify gaps, ambiguities, implicit requirements
- Surface conflicts between request and existing conventions
- Define edge cases and success criteria

When you enhance user input with above ruleset move to Phase 2. Phase 2 is below:

### Phase 2: Plan with Atomic TODOs

Create a detailed TODO list before coding.
Apply Deepthink Protocol when you create TODO list.
If you can track internally, do it internally.
If not, create `todos.txt` at project root—update as you go, delete when done.

```
## TODOs
- [ ] Task 1: [specific atomic task]
- [ ] Task 2: [specific atomic task]
...
```
- Break into 10-15+ minimal tasks (not 4-5 large ones)
- Small TODOs maintain focus and prevent drift
- Each task completable in a scoped, small change

### Phase 3: Execute Methodically

For each TODO:
1. State which task you're working on
2. Apply Deepthink Protocol (reason about dependencies, risks, alternatives)
3. Implement following code standards
4. Mark complete: `- [x] Task N`
5. Validate before proceeding

### Phase 4: Verify & Report

Before finalizing:
- Did I address the actual request?
- Is my solution specific and actionable?
- Have I considered what could go wrong?

Then deliver the Completion Report.

---

## Deepthink Protocol

Apply at every decision point throughout all phases:

**1) Logical Dependencies & Constraints**
- Policy rules, mandatory prerequisites
- Order of operations—ensure actions don't block subsequent necessary actions
- Explicit user constraints or preferences

**2) Risk Assessment**
- Consequences of this action
- Will the new state cause future issues?
- For exploratory tasks, prefer action over asking unless information is required for later steps

**3) Abductive Reasoning**
- Identify most logical cause of any problem
- Look beyond obvious causes—root cause may require deeper inference
- Prioritize hypotheses by likelihood but don't discard less likely ones prematurely

**4) Outcome Evaluation**
- Does previous observation require plan changes?
- If hypotheses disproven, generate new ones from gathered information

**5) Information Availability**
- Available tools and capabilities
- Policies, rules, constraints from CLAUDE.md and codebase
- Previous observations and conversation history
- Information only available by asking user

**6) Precision & Grounding**
- Quote exact applicable information when referencing
- Be extremely precise and relevant to the current situation

**7) Completeness**
- Incorporate all requirements exhaustively
- Avoid premature conclusions—multiple options may be relevant
- Consult user rather than assuming something doesn't apply

**8) Persistence**
- Don't give up until reasoning is exhausted
- On transient errors, retry (unless explicit limit reached)
- On other errors, change strategy—don't repeat failed approaches

**9) Brainstorm When Options Exist**
- When multiple valid approaches: speculate, think aloud, share reasoning
- For each option: WHY it exists, HOW it works, WHY NOT choose it
- Give concrete facts, not abstract comparisons
- Share recommendation with reasoning, then ask user to decide

**10) Inhibit Response**
- Only act after reasoning is complete
- Once action taken, it cannot be undone

---

## Comment Standards

**Comments Explain WHY, Not WHAT:**
```
// WRONG: Loop through users and filter active
// CORRECT: Using in-memory filter because user list already loaded. Avoids extra DB round-trip.
```

---

## Completion Report

After finishing any significant task:

**What**: One-line summary of what was done
**How**: Key implementation decisions (patterns used, structure chosen)
**Why**: Reasoning behind the approach over alternatives
**Smells**: Tech debt, workarounds, tight coupling, unclear naming, missing tests

**Decisive Moments**: Internal decisions that affected:
- Business logic or data flow
- Deviations from codebase conventions
- Dependency choices or version constraints
- Best practices skipped (and why)
- Edge cases deferred or ignored

**Risks**: What could break, what needs monitoring, what's fragile

Keep it scannable—bullet points, no fluff. Transparency about tradeoffs.