PESTEL Strategic Intelligence Skill for OpenClaw

Purpose

Use this skill to run decision-driven macro-environment research with the PESTEL framework and convert raw secondary data into strategic intelligence.

This skill is not for collecting general industry trends. It is for supporting a specific business, product, market, policy, or industry decision with traceable evidence.

This OpenClaw-ready version uses capability-based source routing. It does not require a fixed tool name. The agent should first decide which retrieval or analysis capability is needed, then choose the best available tool in the runtime that satisfies that capability.

Decision -> Missing Information -> Required Capability -> Best Available Tool -> Evidence Check -> Intelligence -> Action

Core Principle

Always begin with the decision, not with search.

Decision -> Missing Information -> PESTEL Map -> Capability Routing -> Targeted Collection -> Evidence Scoring -> Strategic Intelligence

Do not start with broad web discovery. Start from the decision, identify missing information, then route each missing information item to the highest-value available capability.

When to Use

Use this skill when the user needs to research an industry, market, product opportunity, startup idea, investment direction, policy risk, technology trend, or competitive environment.

Typical triggers:

• "Analyze the macro environment for..."
• "Help me research this industry with PESTEL..."
• "What external factors affect this business decision?"
• "Collect industry intelligence for..."
• "Should we enter this market?"
• "Use my database and RSS sources to research..."
• "Monitor industry signals from reliable sources..."

Capability-Based Routing Policy

The skill must describe what capability is required, not which exact tool must be used.

A capability is an abstract function such as:

The agent should inspect available tools in its runtime and choose the tool that best provides the required capability. If multiple tools provide the same capability, choose based on source authority, freshness, user authorization, precision, and ability to return citations or metadata.

Mandatory Source Priority Order

For every research task, follow this source order:

1. User-owned knowledge base / private database
• Local Markdown notes
• Zotero exports
• Project reports
• Previous research database
• Internal CSV / JSON / SQLite records
• User-uploaded documents

1. Configured authoritative feeds

• For China-domestic research, prefer cn_* source groups in config/rss_sources.yml: official statistics, central government/ministry/regulator pages, central state media RSS, major news agency RSS, and reputable domestic finance/technology media feed routes.
• The skill specifies feed capabilities and source requirements only; the runtime chooses the concrete RSS/feed retrieval tool or service.
• If a feed is generated through RSSHub or another feed service, the original publisher and original URL remain the evidence source.
• Government announcements
• Industry regulators
• Reputable financial and technology media
• University or research institution updates
• Standards organizations
• Think tanks and international organizations

1. Official and institutional sources
• Government departments
• Statistics bureaus
• Regulatory bodies
• Legal and standards databases
• International organizations

1. Industry reports and market research
• Consulting firms
• Market research platforms
• Industry associations
• Company annual reports and filings
• Academic papers and conference papers

1. General web discovery fallback
• Use only when higher-priority sources are insufficient
• Must be cross-checked with at least one stronger source when used for important claims

1. LLM internal knowledge
• Use only to explain frameworks or generate hypotheses
• Never use as the sole factual basis for strategic conclusions

If higher-priority sources provide sufficient evidence, do not over-collect from lower-priority sources.

Runtime Tool Selection Rule

Before using any tool, the agent must state the required capability and selection reason.

Use this planning format:

decision_to_support:
key_pestel_dimensions:
missing_information:
required_capability:
source_tier_to_check_first:
tool_selection_criteria:
  - authority
  - freshness
  - relevance
  - user_authorization
  - citation_or_metadata_support
fallback_capability:
cross_validation_plan:
sufficient_evidence_threshold:

The agent should not hard-code a tool name in the plan. It should choose from whatever tools are available in the runtime.

Fallback Rule

If a capability is unavailable, move to the next suitable capability in the source priority order. If no retrieval capability is available, ask the user to provide sources or clearly mark the result as framework-only and not evidence-based.

Required Inputs

Before research, identify these fields. If the user does not provide them, infer a reasonable draft and clearly mark assumptions.

industry:
target_market:
decision_to_support:
time_horizon:
geography:
business_type:
known_constraints:
preferred_source_types:
private_knowledge_scope:
feed_scope:
output_depth:

PESTEL Missing Information Map

Break the decision into missing information questions.

Each missing information item must include:

id:
pestel_dimension:
research_question:
decision_relevance:
source_tier_to_check_first:
required_capability:
expected_evidence:
priority:
validation_requirement:

Evidence Scoring

Score each evidence item from 0 to 25.

High-impact recommendations require at least one strong evidence item or two independent medium evidence items. Legal and policy claims should be verified with official or regulatory evidence whenever possible.

Information-to-Intelligence Conversion

Do not stop at summarizing information. Convert information into intelligence.

For every important finding, produce:

raw_information:
source_type:
source_tier:
evidence_score:
limitation:
strategic_meaning:
decision_implication:
recommended_action:
confidence_level:

Example logic:

Information: A city released a pilot policy supporting autonomous delivery in semi-closed campuses.
Intelligence: This lowers regulatory uncertainty for campus-first MVP testing.
Action: Prioritize pilot cities with explicit smart logistics or robotics policy support.

Required Output Structure

For full tasks, output the following sections:

1. Decision Context
2. Assumptions and Constraints
3. PESTEL Missing Information Map
4. Capability Routing Plan
5. Evidence Scorecard
6. PESTEL Intelligence Matrix
7. Source Log
8. Strategic Options
9. Recommended Action
10. Remaining Unknowns and Next Data Collection Plan

For quick tasks, output:

1. Decision being supported
2. Top 3 missing information items
3. Source/capability plan
4. Key evidence found
5. Strategic implication
6. Next action

Capability Routing Plan Template

Source Log Template

Use capability names in the Access Capability column, not specific tool names.

Guardrails

• Do not produce strategic conclusions from a single weak or low-authority source.
• Do not treat news signals as verified facts unless corroborated.
• Do not use LLM memory as factual evidence.
• Do not hide uncertainty. Mark assumptions, limitations, and confidence.
• Do not over-search if enough high-priority evidence already exists.
• Do not hard-code tool names; route by capability and let the runtime choose the implementation.

Suggested Skill Pack Files

SKILL.md
README.md
config/source_policy.yml
config/rss_sources.yml
config/evidence_scoring.yml
capabilities/private_knowledge_retrieval.md
capabilities/authoritative_feed_monitoring.md
capabilities/evidence_assessment.md
templates/pestel_brief_template.md
templates/intelligence_matrix.md
templates/search_task_template.md
templates/source_log.csv
templates/openclaw/retrieval_plan.md
templates/openclaw/source_router_task.md
templates/openclaw/evidence_scorecard.md
templates/openclaw/openclaw_prompt.md
examples/quadruped_robot_delivery_example.md

One-Shot Prompt Template

Act as my PESTEL Strategic Intelligence Agent.

Decision to support: [decision]
Industry: [industry]
Geography: [geography]
Time horizon: [time horizon]
Available source categories: [private knowledge, authoritative feeds, official sources, reports, academic sources, general web]

Rules:
1. Start from the decision.
2. Identify missing information using PESTEL.
3. For each missing information item, choose the required capability first, not a fixed tool.
4. Let the runtime select the best available tool for that capability.
5. Prioritize user-owned knowledge and authoritative feeds before general web discovery.
6. Verify legal, policy, statistical, and market-size claims with high-authority sources when possible.
7. Score evidence before making recommendations.
8. Convert raw information into strategic implications and recommended actions.

Output:
- Decision Context
- Capability Routing Plan
- PESTEL Intelligence Matrix
- Evidence Scorecard
- Source Log
- Strategic Options
- Recommended Next Step