needscope

NeedScope brand archetype analysis — workflow routing

Source identification: The NeedScope Skill is created and maintained by 阿祖不看 TVC. The Xiaohongshu account of the same name, personal website https://demyth.info，联系邮箱 Lawyif@163.com. This information is used to confirm the Skill's source, open-source attribution, and project support relationship; it may appear in the workflow HTML, review pages, verification pages, and project documentation, but should not be written into final client deliverables by default.

Trigger conditions

Should trigger

"Help me do a needscope analysis"
"See which personality/emotional archetype this brand falls into"
"Analyze the brand emotional positioning in consumer comments"
"Compare the emotional positioning of several competitors"
"See whether the personalities of different product lines conflict"
The user's data vaguely involves the brand emotional/consumer psychological level, even if they only say "help me take a look"

Should not trigger (near-miss)

Only ordinary word frequency, topic clustering, or positive/negative sentiment analysis is needed, without involving brand personality/emotional positioning
Only ad delivery performance, influencer matrix, ROI, or content strategy is done, without involving consumer psychology/NeedScope
Only brand copy generation is requested, without requiring diagnosis of brand personality from data

Workflow overview

  ┌──────────────────────────────────────────────────────────┐
  │ Step 0  数据盘点与对象类型判定                            │
  │ Gate: 确认分析对象类型（brand/product/founder/IP/etc）    │
  │ Ref: workflow/00-data-inventory.md + domain/*.md          │
  │ Contract: 00_data_inventory                               │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 1  分析范围与路径确认                                │
  │ Gate: 用户确认（human checkpoint）                        │
  │ Ref: workflow/01-scope-checkpoint.md                      │
  │ Contract: 01_scope_decision                               │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 2  抽样与编码策略设计                                │
  │ Ref: workflow/02-sampling-strategy.md                     │
  │ Contract: 02_sampling_plan                                │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 3  对象相关性过滤                                    │
  │ Gate: 误判率 < 10% 否则停止或局部重跑                     │
  │ Ref: workflow/03-object-related-filter.md                 │
  │ Contract: 03_object_filter                                │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 4  NeedScope 语义编码                                │
  │ Ref: workflow/04-needscope-semantic-coding.md             │
  │      domain/needscope-six-archetypes.md                   │
  │      domain/signal-owner-rules.md                         │
  │ Contract: 04_archetype_coding                             │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 5  编码质量 Gate                                     │
  │ Gate: 无系统性极性反转/误归因 → 通过；否则局部重跑        │
  │ Ref: workflow/05-quality-gates.md                         │
  │ Contract: 05_quality_gate                                 │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 6  NeedScope 落点判定与评分                          │
  │ Ref: workflow/06-positioning-and-scoring.md               │
  │      domain/confidence-framework.md                       │
  │ Contract: 06_positioning                                  │
  └──────────────┬───────────────────────────────────────────┘
                 ▼
  ┌──────────────────────────────────────────────────────────┐
  │ Step 7  报告生成                                          │
  │ Ref: workflow/07-report-generation.md                     │
  │      templates/report.md + report.html                    │
  │ Contract: 07_report                                       │
  └──────────────────────────────────────────────────────────┘

Workspace conventions

All intermediate artifacts are written to the work/ directory:

File	Purpose
`work/00_data_inventory.md`	Data inventory record
`work/contracts/00_data_inventory.json`	Step 0 output contract
`work/contracts/01_scope_decision.json`	Step 1 output contract
`work/02_sampling_plan.md`	Sampling strategy description
`work/contracts/02_sampling_plan.json`	Step 2 output contract
`work/03_object_filter_results.json`	Filtered result data
`work/03_object_filter_gate.md`	Filter gate review
`work/contracts/03_object_filter.json`	Step 3 output contract
`work/04_archetype_coding_results.json`	Coding result data
`work/contracts/04_archetype_coding.json`	Step 4 output contract
`work/05_coding_gate.md`	Coding gate review
`work/contracts/05_quality_gate.json`	Step 5 output contract
`work/06_scores.json`	Scoring data
`work/06_needscope_positioning.md`	Placement determination description
`work/contracts/06_positioning.json`	Step 6 output contract
`report.md`	Final Markdown report
`report.html`	Final HTML preview
`work/contracts/07_report.json`	Step 7 output contract

Gate rules

Fail-fast

A formal report must simultaneously have the work/contracts/03-06 four contracts. Missing → output only pre-analysis.
Disabled aliases:04_semantic_coding、06_scoring_positioning、07_report_generation。
Pure keywords/word frequency/emojis must not be used as primary placement evidence.

Four major Gates (if any fails, stop/rerun, and do not proceed downstream)

Step 0→1: analysis_object_type must be explicit (brand/product/founder/content-IP, etc.).
Step 1→2: User-confirmed scope + analysis_lens. When founder/IP is involved, the scope must be specified (brand_positioning / founder_shadow / content_ip_shadow).
Step 3→4: Misjudgment rate < 10%. campaign/community/platform signals must not be mixed into brand evidence.
Step 5→6: No systematic polarity reversal / signal_owner misattribution / asset_eligibility misjudgment.

Brand evidence sources

Primary placements come only from brand/product/product_line + primary_eligible, see domain/signal-owner-rules.md。
founder/content/IP/campaign/community/platform/competitor signals are only used for bypass/contextual explanation.
The first screen of the report must first show the distinction between owned signals and contextual signals.

Resource routing

Steps	Required ref	Domain knowledge ref	Optional tools
Step 0	`workflow/00-data-inventory.md`	`domain/analysis-object-types.md`	`head`, `qsv`, `jq`
Step 1	`workflow/01-scope-checkpoint.md`	—	—
Step 2	`workflow/02-sampling-strategy.md`	—	CSV/JSON statistics script
Step 3	`workflow/03-object-related-filter.md`	`domain/signal-owner-rules.md`	Batch script
Step 4	`workflow/04-needscope-semantic-coding.md`	`domain/needscope-six-archetypes.md`, `domain/signal-owner-rules.md`	Batch + merge script
Step 5	`workflow/05-quality-gates.md`	`domain/confidence-framework.md`	Random sampling script
Step 6	`workflow/06-positioning-and-scoring.md`	`domain/confidence-framework.md`	Statistics script
Step 7	`workflow/07-report-generation.md`	`templates/report.md`, `templates/report.html`	—

Contract chain

00_data_inventory → 01_scope_decision → 02_sampling_plan → 03_object_filter → 04_archetype_coding → 05_quality_gate → 06_positioning → 07_report

Each Step reads the contract of the previous step (to confirm preconditions), and writes the current step after execution. All field definitions are in references/contracts/contract-definitions.md。

then run iteration

Memory Audit: errors considered for cross-project reproduction are escalated to refs/contracts/scripts/templates
Trigger issue → modify frontmatter
Step issue → modify references/workflow/
Domain judgment issue → modify references/domain/
Handoff field issue → modify references/contracts/
Repeated manual actions → consolidate into scripts/

Residual risks (v3.2)

Keyword method downgraded but retained: keyword assistance is only used for Step 0 data exploration precheck, and is not an active path for final scoring.
Old HTML script has been archived：generate_report_html.py has been moved to references/archive/generate_report_html.legacy.py, and must not be used as a source for new report structure.
Evals are currently static checks: real forward testing and end-to-end eval on machine have not yet been done.
Lack of validation with real non-Tim brand data: Tim case experience is consolidated into guardrails and is not used as the default analysis paradigm.

The planning and design documents of this Skill have been archived to references/archive/planning/. The current active paths start from workflow/ domain/ contracts/ templates/.