An AI-assisted, human-controlled workflow for discovering, evaluating, preparing, validating, and tracking enterprise policy applications.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install policy-application-advisor
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install policy-application-advisor using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Policy Application Advisor?
Policy Application Advisor is one of the Openclaw Skills designed to support the complete policy and subsidy application lifecycle for companies, especially businesses operating in industrial parks or economic zones. It connects policy radar, eligibility assessment, document preparation, form-filling assistance, human review, and approval progress tracking in a structured six-step workflow.
The skill provides analysis, estimates, draft content, consistency checks, reminders, and actionable next steps while keeping decision-making and external submission with the user. It uses strict BPM stage gates, a clarification-first grilling mechanism, read-only memory, source freshness checks, and a mandatory human confirmation gate to reduce rework without autonomously submitting applications or sending sensitive information.
Policy Application Advisor Use Cases
- Discover applicable national, provincial, municipal, district, or industrial-park policies based on an enterprise profile.
- Match company qualifications against high-tech enterprise, specialized and innovative enterprise, R&D subsidy, rent subsidy, talent subsidy, ISO, software copyright, and patent programs.
- Evaluate whether a policy application is worthwhile by comparing estimated subsidy value with application effort and cost.
- Generate draft application forms, feasibility reports, financial estimates, and policy-specific materials for user review.
- Identify missing documents, inconsistent figures, expired enterprise data, and common rejection risks before submission.
- Map confirmed materials to application form fields and check required fields and formatting rules.
- Track acceptance, preliminary review, correction, public notice, and payment milestones.
- Trigger deadline warnings at 15, 7, and 3 days before a policy window closes.
- Prepare correction recommendations when authorities request supplementary materials.
- Archive completed application projects and feed confirmed history into future policy matching.
How Policy Application Advisor Works
- Clarify the request when necessary. If the objective, policy name, or draft type is ambiguous, the skill performs one structured clarification round. It first checks available enterprise, policy, and historical data, asks only for unresolved dependencies, and provides recommended defaults.
- Run policy radar. The skill compares the enterprise profile with the policy database, calculates matching and compliance signals, identifies gaps, estimates potential subsidies, records policy windows, and highlights upcoming deadlines.
- Evaluate project feasibility. For a selected policy, it reviews qualification status, estimates input and output, compares expected subsidy value with application cost, and recommends applying, postponing, or not applying with supporting reasons. The user makes the final decision.
- Prepare application materials. Based on policy guidance, enterprise data, historical applications, and templates, it creates draft application books, feasibility report drafts, financial calculation tables, and a missing-material checklist. Every generated item is marked as pending user confirmation.
- Assist with form completion and validation. Confirmed materials are mapped to application fields. The skill checks required fields, formats, logical consistency, attachment status, and historical rejection risks. Failed checks block progression to confirmation.
- Apply the human confirmation gate. The user reviews the final checklist, verifies material authenticity and completeness, and explicitly confirms before any submission-related action. Without confirmation, no submission action is produced.
- Track delivery and outcomes. The skill analyzes approval nodes, detects status changes, reminds users about new milestones, interprets correction requests, and recommends supplementary materials. After completion, the project is archived and its confirmed history can support future matching.
Policy Application Advisor Setup
Installation and configuration
This SKILL.md does not define a standalone package, dependency, or CLI installer. Deploy Policy Application Advisor through the OpenClaw skill registry or the host workspace mechanism used by your Openclaw Skills environment, then load the skill metadata and workflow instructions.
- Add the skill package to the OpenClaw skills workspace and enable the
policy-application-advisorskill. - Provide read access to the policy database, including policy source, issuing level, publication date, deadline, eligibility criteria, application guide, and subsidy standards.
- Provide read access to the enterprise profile, qualification ledger, material template library, and historical application records.
- Configure the enterprise profile fields used for matching, such as industry, company scale, revenue, headcount, R&D investment, intellectual property, certifications, and enterprise identity.
- Connect or enable
data-security-verifierfor source classification, data freshness checks, field validation, and de-identification before output. - Configure the policy-window warning thresholds to 15, 7, and 3 days before the deadline.
- Confirm that generated drafts are presented for review rather than automatically persisted, and that external submission, messaging, API calls, and third-party writes remain disabled until the user acts.
- Test each BPM gate with a sample policy project: eligibility assessment, material preparation, form validation, human confirmation, and progress tracking.
No external submission command is supplied by this skill. If the host requires CLI-based registration, use the host platform's documented skill-management command rather than assuming a package-specific installer.
# No standalone installation command is defined in this SKILL.md.
# Register and enable the skill using your OpenClaw host's documented workflow.
Policy Application Advisor Data Schema & Taxonomy
Core project model
Each policy opportunity is represented as one application project containing stages, gate results, evidence, user decisions, and delivery status.
| Data domain | Typical fields | Primary use |
|---|---|---|
| Policy database | Policy name, administrative level, source, publication date, deadline, eligibility rules, subsidy standard, application guide | Matching, compliance analysis, deadline monitoring, and subsidy estimation |
| Enterprise profile | Industry, scale, revenue, headcount, R&D investment, intellectual property, certifications, and qualification status | Eligibility matching and gap analysis |
| Qualification ledger | High-tech status, specialized and innovative status, ISO, software copyrights, patents, validity dates | Self-assessment and expiry warnings |
| Material template library | Draft type, required sections, field rules, attachment requirements, formatting constraints | Draft generation and form mapping |
| Application history | Historical and current projects, decisions, rejection risks, submitted materials, outcomes, and confirmed dates | Reuse, rejection-risk warnings, project management, and future matching |
| Project workflow | Policy target, stage, gate status, missing inputs, owner, user decision, next action, and deadline | BPM control and dashboard visibility |
| Approval timeline | Acceptance, preliminary review, correction, public notice, payment, status timestamp, and reminder state | Progress tracking and milestone notifications |
Generated outputs
- Policy match results with compliance status, qualification gaps, estimated subsidy, source details, and application window.
- Feasibility assessment with apply, postpone, or do-not-apply recommendation and rationale.
- Material checklist and acquisition recommendations.
- Draft application book, feasibility report, and financial calculation table.
- Form-field mapping, required-field validation results, format checks, logical consistency findings, attachment checks, and rejection-risk alerts.
- Correction response recommendations and milestone reminders.
- A final human confirmation checklist.
Data governance and metadata taxonomy
- Source metadata: source name, administrative level, publication date, effective date, deadline, and freshness status.
- Evidence metadata: enterprise data field,
data_as_oftimestamp, verification state, and whether the value was user-confirmed. - Confidence labels: estimated subsidy values are marked as estimates and must defer to official figures.
- Privacy labels: company names, amounts, and certificate numbers may be masked as
XXin generated outputs. - Workflow labels: draft, pending user confirmation, blocked by missing materials, ready for review, confirmed, submitted by user, correction required, public notice, paid, and archived.
- Memory policy: read-only by default for this skill. Drafts are presented to the user; the skill does not autonomously write project state, modify source databases, or send enterprise financial or tax originals to third parties.
Policy Application Advisor Advanced Features
- Clarification-first grilling: Detects vague objectives, missing policy names, missing draft types, and high-impact ambiguous choices before execution.
- Frontier-based questioning: Collects all currently resolvable required questions in one numbered round, includes recommended defaults, and avoids repetitive back-and-forth.
- Event-driven workflow: Responds to policy database updates, application-window openings, enterprise-profile changes, approval-node changes, correction requests, public notices, and payment events.
- BPM stage gates: Prevents material generation when key conditions fail, blocks form filling when required materials are missing, and blocks confirmation when required fields fail validation.
- Human-in-the-loop control: Enforces explicit user review and confirmation before submission-related actions; the AI never substitutes for the user's decision or approval.
- Read-only memory architecture: Reads policy, enterprise, template, and historical data while avoiding autonomous persistence, source modification, or third-party writes.
- Data-security-verifier integration: Supports source grading, freshness validation,
data_as_ofchecks, field parsing errors, de-identification, and safe output handling. - Deadline intelligence: Uses 15-, 7-, and 3-day warning levels and suppresses application alerts for expired policies until the policy database is refreshed.
- Rejection-risk analysis: Uses historical rejection records to identify common application pitfalls and recommend corrective actions.
- Approval-chain monitoring: Tracks acceptance, preliminary review, correction, public notice, and payment stages, with reminders for new milestones.
- Post-project learning: Archives completed projects, updates the qualification ledger through user-controlled actions, and uses confirmed application history to improve subsequent policy matching.
- Multi-source policy taxonomy: Organizes opportunities across national, provincial, municipal, district, and industrial-park levels while retaining source and validity metadata.
SKILL.md
Loading
Related Openclaw Skills
Dimensional-Breaking Interactive Playbook | AI-HIVE Original Workflow
A rights-aware AI-HIVE workflow for planning, generating, reviewing, and delivering original animation and game-oriented image and video content.
Agorism in the 21st Century: Volume 2 Knowledge Base
A focused research skill for exploring the philosophical frameworks, essays, terminology, and tensions in Agorism in the 21st Century: Volume 2.
Everything Is a Battlefield | AI-HIVE Original Content Workflow
An AI-HIVE MCP workflow for planning, generating, reviewing, and delivering original animation and game content without reproducing protected IP.
Smart Medical Device IoMT Compliance
A compliance writing assistant that maps connected medical devices across hardware, software, sensors, apps, and cloud services to NMPA, FDA, and MDR requirements.
Helena Marketing Agent
Helena connects OpenClaw to Enrich Labs so an AI marketing agent can research, plan, create, schedule, publish, and analyze cross-channel brand marketing.
txt.by Public Agent Messaging
txt.by enables AI agents to publish, discover, and coordinate through public Markdown messages, threads, profiles, and inboxes.
Featured*
