aggarwalgauri05/openclaw/workspace-policy/skills/intent-validator
openclaw/workspace-policy/skills/intent-validator/SKILL.md
# intent-validator Skill This skill classifies incoming citizen messages into one of the following categories: - POLICY_QUERY - ELIGIBILITY_CHECK - BENEFIT_MATCH - APPLICATION_GUIDE - OUT_OF_SCOPE It then determines the most appropriate source for retrieving information and logs this decision. ## Functionality 1. **Intent Classification**: Identifies the user's intent using keyword matching. 2. **Source Routing**: Based on keywords in the message, it determines the information source:
development
Updated Apr 3, 2026
$ install --global
skillsauthnpx skillsauth add aggarwalgauri05/citizenclaw openclaw/workspace-policy/skills/intent-validatorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 3, 2026, 10:25 PM58.5s2 files scanned
SKILL.md
intent-validator Skill
This skill classifies incoming citizen messages into one of the following categories:
- POLICY_QUERY
- ELIGIBILITY_CHECK
- BENEFIT_MATCH
- APPLICATION_GUIDE
- OUT_OF_SCOPE
It then determines the most appropriate source for retrieving information and logs this decision.
Functionality
- Intent Classification: Identifies the user's intent using keyword matching.
- Source Routing: Based on keywords in the message, it determines the information source:
SOURCE_LIVE: For queries about current, latest, or today's information (e.g., 'current', 'latest', 'today'). This suggests fetching data from live sources like government portals.SOURCE_NEWS: For queries related to recent updates or news concerning schemes (e.g., 'recent', 'new scheme', 'update', 'change'). This indicates a need to check news or announcements.SOURCE_PDF: For requests seeking explicit documentation or rules (e.g., 'official rules', 'what does the act say', 'guidelines', 'document'). This implies looking for official PDF documents.SOURCE_MEMORY: For general questions or existing knowledge recall where no specific external source is indicated (default).
- Response Formatting: Every response starts with:
[INTENT: <classified intent>] [SOURCE: <determined source>] <your helpful response> - OUT_OF_SCOPE Handling: If a message is classified as
OUT_OF_SCOPE, the skill immediately responds with "I can only help with government schemes and benefits." and stops further processing. - Logging: The intent classification and source decision are logged before any actions are taken.
Overriding Rules
This skill enforces the following rules, which override all other instructions:
- Rule 1: ALWAYS classify intent before proceeding.
- Rule 2: ALWAYS start EVERY response with
[INTENT: <classified intent>]. - Rule 3: ALWAYS log the source decision as
[SOURCE: <type>]on the second line of the response. - Rule 4: OUT_OF_SCOPE queries are BLOCKED immediately.
Related Skills
aggarwalgauri05/openclaw/workspace-policy/skills/report-writer
development
VerifiedTrustedCommunity
# report-writer Skill This skill generates a personalized benefits report for a citizen based on their profile information. ## Activation The skill can be invoked manually (e.g., via a command line) or integrated into a workflow after intent classification (e.g., when the user asks for a summary of applicable schemes). ## Command‑line interface ```bash python write_report.py \ --citizen-name <NAME> \ --age <AGE> \ --income <INCOME> \ --occupation <OCCUPATION> \ --state <
SKILL.mdUpdated Apr 3, 2026
aggarwalgauri05/openclaw/workspace-policy/skills/policy-interpreter
development
VerifiedTrustedCommunity
# policy-interpreter Skill This skill interprets government schemes and benefits when a user's intent is identified as `POLICY_QUERY` or `ELIGIBILITY_CHECK`. ## Functionality When activated for `POLICY_QUERY` or `ELIGIBILITY_CHECK` intents: 1. **Provide Plain-Language Explanations**: It fetches information about the specified government scheme and presents it in an easy-to-understand manner. 2. **List Eligibility Criteria**: It extracts and lists the eligibility requirements as numbered st
SKILL.mdUpdated Apr 3, 2026
aggarwalgauri05/openclaw/workspace-policy/skills/pdf-reader
development
VerifiedTrustedCommunity
# pdf-reader Skill This skill provides secure access to reading PDF documents from an approved directory. ## Functionality 1. **Approved Directory**: Documents can only be accessed if they reside within the `~/.openclaw/workspaces/default/docs/govt-policies` directory. 2. **File Access Control**: The script enforces strict access control. Any attempt to read a file outside this directory will be blocked. 3. **Argument Handling**: * `--file <PATH>`: Specifies the PDF file to read.
SKILL.mdUpdated Apr 3, 2026
aggarwalgauri05/openclaw/workspace-policy/skills/gov-fetcher
development
VerifiedTrustedCommunity
# gov-fetcher Skill This skill allows fetching content from specific, approved government domains. It acts as a safeguard to ensure that external data retrieval is limited to trusted sources. ## Functionality 1. **Domain Whitelisting**: The skill maintains a strict list of approved domains (all ending in `.gov.in`). Any attempt to fetch content from a domain not on this whitelist will be blocked. 2. **URL and Query Parameters**: It accepts a `--url` argument specifying the target web addre
SKILL.mdUpdated Apr 3, 2026