OPC 案例调研

Systematic public-information research for OPC, super-individual, creator IP, and one-person business cases, with Chinese outputs focused on content strategy...

Security Scan

This is an instruction-only research skill whose requested capabilities, instructions, and file assets align with its stated purpose and do not ask for secrets or install code.

Purpose & Capability

The name and description (OPC case research focused on creator/one-person-business cases) match the SKILL.md and included templates/references. No unrelated binaries, env vars, or external credentials are requested.

Instruction Scope

Runtime instructions are limited to public-information research, building source maps, timelines, and evidence tables; they explicitly forbid digging for private data and emphasize separating facts/inferences/unknowns. The guidance does not instruct the agent to read system files, exfiltrate environment variables, or contact unexpected endpoints.

Install Mechanism

No install spec and no code files (instruction-only). Nothing will be downloaded or written to disk by the skill itself.

Credentials

The skill requires no environment variables, credentials, or config paths. Requested assets are local templates and reference docs consistent with the stated function.

Persistence & Privilege

Flags are default: always:false and normal autonomous invocation allowed. The skill does not request permanent presence or system-level configuration changes.

Assessment

This skill is instruction-only and coherent with its stated purpose: it uses public sources, provides templates, and asks for no credentials or installs. Before installing, note that (1) outputs default to Chinese—confirm language if you need another; (2) the skill emphasizes using public information and avoiding private data—do not prompt it to research non-public or sensitive personal details; and (3) although the skill itself is benign, any research report could contain unverified inferences—ask for clear source links and the evidence snapshot if you require traceability or legal caution.

OPC Case Research

Overview

Use this skill to research a specific person, creator brand, or one-person business case as an "OPC / super-individual / creator IP" example.

Default to:

• public information only
• Chinese output
• single-case research
• emphasis on content strategy, IP positioning, and business model
• explicit separation of facts, inferences, and unknowns

Quick Start

1. Confirm the target and scope. If the user did not specify depth, use the standard mode.
2. Build a base profile and source map before writing conclusions.
3. Build the timeline before interpreting strategy or causality.
4. Analyze content system, IP positioning, channels, and business model together.
5. End with replicability analysis instead of generic praise.

If the request is only a fast initial screen, do not load every reference file. Read only what is needed.

Default Inputs

Treat these as the input contract:

• Required: target person, brand, or case name
• Optional: focus areas such as content, IP, business model, channels, key decisions, or methods
• Optional: depth
  • quick: short case brief
  • standard: structured research output
  • deep: standard output plus evidence tables, samples, and estimates
• Optional: time range
• Optional: comparison target

If the user does not specify otherwise, assume:

• use public information
• write in Chinese
• prioritize content strategy, IP building, and monetization
• analyze one case at a time
• include a minimum evidence slice in standard and deep outputs

Workflow

1. Scope the Request

Resolve only the ambiguity that blocks the work. If the target is clearly identifiable, start immediately.

Set the working mode:

• quick: decide whether the case is worth deeper study
• standard: produce the default structured case study
• deep: add evidence tables, sampling, and cautious estimates

2. Build the Base Profile and Source Map

Collect:

• identity labels
• platform presence
• official pages or landing pages
• visible business entry points
• likely primary and secondary sources

Prefer source types in this order:

• A: first-party statements, official pages, original program or event pages
• B: mainstream media, platform profile pages, databases, partner pages
• C: reposts, forums, summaries, comments, only for leads

Read references/search-playbook.md when constructing search queries, sampling plans, or source maps.

3. Build the Timeline First

Do not jump straight into opinions. Build an event sequence first.

Minimum expectations:

• quick: 5+ meaningful nodes
• standard: 10+ nodes
• deep: 10-20 nodes plus notes on why each node matters

Track:

• role shifts
• platform shifts
• format shifts
• commercialization upgrades
• major public launches or collaborations

4. Analyze the Four Core Layers

Analyze these together, not in isolation:

1. Identity and positioning
2. Content system and channel strategy
3. Business model and monetization structure
4. Key decisions, turning points, and path evolution

Read references/research-standard.md for the full checklist and evaluation rules.

5. Classify Evidence Carefully

Every important point should be marked as one of:

• fact
• inference
• unknown

When in doubt, downgrade confidence instead of overstating certainty.

Use cautious language for:

• revenue structure
• team size
• conversion assumptions
• audience profile assumptions
• operational scale

Read references/evidence-schema.md when building evidence tables or appendices.

6. Build a Minimum Evidence Slice

Even in standard mode, do not leave evidence fully implicit.

Default to a minimum evidence slice with 6-12 rows or bullet-equivalents that support the most important claims across:

• identity or self-positioning
• timeline
• content or IP
• channels
• business model

If the user did not ask for tables, the evidence slice can be a compact appendix or a short "evidence snapshot" section instead of a full spreadsheet.

7. Shrink Claims When Public Information Is Thin

When evidence is weak, incomplete, or highly indirect:

• narrow the scope instead of compensating with confident writing
• separate confirmed facts, best-effort inferences, and unknowns
• reduce the depth of business-model claims first
• keep a to verify next list if the case is still worth researching

A smaller but more reliable output is better than a complete-looking report built on speculation.

8. Produce the Right Deliverable

Choose the smallest deliverable that still answers the user.

For quick, produce:

• one-line positioning
• identity tags
• main platforms
• visible monetization entry points
• 3-5 reasons the case matters
• open questions

For standard, produce:

• case brief
• source map summary
• timeline
• content and IP analysis
• channel analysis
• business model analysis
• evidence snapshot or minimum evidence slice
• replicability analysis
• limitations and unknowns

For deep, add:

• full evidence table
• content sampling table
• business model table
• estimate disclosures when needed

Read references/report-template.md before drafting a full report. Use the copyable templates in assets/ when the output should become a reusable document or table.

Output Requirements

Always aim for:

• Chinese writing
• research tone, not fan tone
• visible structure
• evidence awareness
• at least a small evidence trail for the main claims
• explicit dates or time ranges when claims depend on time
• direct linkage between content strategy, channel choice, and business model
• actionable takeaways for people studying super-individual paths

Hard Rules

Do not:

• use private or non-public information
• present unverified claims as facts
• invent revenue, team, or deal details
• turn gossip into business analysis
• hide uncertainty behind confident phrasing

When information is weak, say so clearly and narrow the claim.

Resource Map

• references/research-standard.md
  • Read for the full SOP, evidence rules, deliverables, quality bar, and uncertainty handling.
• references/search-playbook.md
  • Read for search query patterns, source mapping, verification rules, and sampling.
• references/evidence-schema.md
  • Read for field definitions for evidence, timeline, content samples, and business model tables.
• references/report-template.md
  • Read when drafting a standard long-form report.
• assets/case-brief-template.md
  • Use when creating a short case brief.
• assets/research-report-template.md
  • Use when creating a full report draft.
• assets/evidence-table.csv
  • Use when creating the evidence table.
• assets/timeline-table.csv
  • Use when creating the timeline table.
• assets/content-sample-table.csv
  • Use when creating the content sampling table.
• assets/business-model-table.csv
  • Use when creating the business model table.

Not for This Skill

This skill is not for:

• celebrity gossip
• private intelligence gathering
• legal, financial, or investment due diligence
• unsupported claims about real income or private operations